4 Ethekwini Design Manualmay 2008

ENGINEERING UNIT
COASTAL STORMWATER & CATCHMENT MANAGEMENT DEPARTMENT
DESIGN MANUAL:
GUIDELINES AND POLICY FOR THE DESIGN OF STORMWATER
DRAINAGE AND STORMWATER MANAGEMENT SYSTEMS
(Vade Mecum)
Issued by :
R. KASSERCHUN PrENG.
DEPUTY HEAD : CSCM
MUNICIPAL CENTRE
166 OLD FORT ROAD
DURBAN
4001
Revision Date: May 2008 1

ETHEKWINI MUNICIPALITY
DESIGN MANUAL: GUIDELINES AND POLICY FOR THE DESIGN OF

STORMWATER DRAINAGE AND STORMWATER MANAGEMENT
SYSTEMS
INTRODUCTION
These guidelines and policies are applicable to the design of minor stormwater reticulation
and collection systems and for the management and control of stormwater runoff from sites
by means of soakpits and attenuation tanks and runoff discharge controls.
These guidelines include recommendations for the sizing and design of stormwater soakpits,
attenuation structures/ponds, outlet discharge controls/overflows, kerb inlets, manholes, road
edge channels, watercourses, underground pipelines and small channels. These must be
designed to effectively collect, control and convey run-off from storms to larger or major
drainage systems.
The design of all major stormwater systems (whether culverts, pipes, canals, including road
crossings, bridges etc.), for catchments greater than 1 km where hydraulic analysis of the
waterway is necessary, should be carried out by checking by a registered professional
engineer proficient in this field for approval by Coastal, Stormwater & Catchment
Management Unit.
Limited information related to the drainage and storage facilities to be used in a major
drainage system for the control of floods has been included in these guidelines.
These guidelines should be read in conjunction with the Stormwater management section of
the Guidelines for the provision of Engineering Services and Amenities in Residential
Township Development (the Red book) issued by the National Housing Board.
Rail and highway crossings and jacked pipes normally require special measures and should
be referred to the Deputy Head: Coastal Stormwater and Catchment Management.

CONTENTS
1 STORMWATER MANAGEMENT POLICY page 5
1.1 Provision of soakpits

1.2 Provision of attenuation and other storage structures
1.3 Brief outline of minimum information required in a stormwater management plan
2 THE RATIONAL METHOD page 10
2.1 Restriction on usage

2.2 Area of catchment 'A'
2.3 Time of concentration 'Tc'
2.4 Run-off coefficient 'C'
2.5 Rainfall intensity 'I'
3 DESIGN OF A MINOR SYSTEM page 14
3.1 Critical points

3.2 Rainfall intensity at critical points
3.3 Minimum diameter
3.4 Minimum velocity and gradient
3.5 Materials
3.6 Anchor blocks
3.7 Curved alignment
3.8 Servitudes
3.9 Layout of stormwater sewers in servitudes
3.10 Layout of stormwater sewers in road reserves
3.11 Manholes
3.12 Manhole covers
3.13 Benching in manholes
3.14 Channels
3.15 Minimum cover
3.16 Bedding
3.17 Invert levels at manholes
3.18 Inlets
4 SELECTION OF CLASS OF PIPE page 20
5 PARTIAL FLOW IN PIPES page 20
6 SUBSOIL DRAINS page 20
7 STORAGE PONDS page 20
8 DESIGN OF SHORT LENGTH CULVERTS UNDER EMBANKMENTS page 21
9 CALCULATING TIME OF CONCENTRATION page 21
10 EXAMPLES OF CALCULATING THE RUNOFF COEFFICIENT C page 22
11 METHOD OF STORMWATER RETICULATION DESIGN page 24
12 GENERAL METHODOLOGY page 24
13 WORKED EXAMPLE page 27

14 DESIGN OF STORMWATER KERB INLETS page 33
15 CALCULATIONS OF INVERT LEVELS AT MANHOLES page 34
LIST OF APPENDICES page 35
PART R Stormwater Disposal SABS 0400 - 1990 page 36

Hydraulic elements of circular sections page 38
Bend losses page 39
Details of anchor blocks page 40
Discharge through box culverts : inlet control page 41
Discharge through pipe culverts : inlet control page 42
Kerb Inlet chart 1: cross fall 2% page 43
Kerb Inlet chart 2: cross fall 2,5% page 44
Kerb Inlet chart 3 & 4 : cross fall 4% and 6% page 45
Pipe bedding details page 46
Protection of pipes at reduced depths of cover page 47
Recommended layout of services - 9.5 metre reserve page
48
Recommended layout of services - 12 metre reserve page 49

1 STORMWATER MANAGEMENT POLICY
Municipal stormwater availability
In many instances erven do not have access or connections to municipal stormwater. In

such cases the onus is placed on the developer to manage the excess stormwater resulting
from any hardening of the site area. This management can take the form of soakpits if the
soil and geological profiles allow or attenuation facilities such as tanks, ponds or areas
designed to retain water. Where possible the emphasis should be placed on improving the
potential for groundwater infiltration.
Municipal infrastructure generally provided stormwater systems designed on the basis that
not more than 40% of the area of residential properties would be hardened. As such, any
development in such areas in excess of a 40% limitation naturally implies that the developer
must be held responsible to manage the excess runoff from such a site for the proportion of
hardening in excess of 40%.
Where sites have historically disposed of stormwater into soakpits on site, this method of
disposal must be maintained. Notwithstanding, soakpits cannot function effectively forever
(life spans are limited to 5 to 15 years depending on the design and maintenance). The
obligation is thus placed on the owner to maintain and upgrade the existing soakpit capacity
where any further development or hardening of the site occurs.
These responsibilities are necessary and due. Changing conditions and trends have lead to
increased development densification. Increasingly, paving is replacing higher maintenance
lawns and garden areas and a process of maximizing land usage to supplement incomes.
These changing trends in residential and semi-urban areas have seen rezoning and new
sectional type developments resulting in higher rates and volumes of runoff.
Therefore, In accordance with the National Building Regulations PART R Stormwater

Disposal SABS 0400 1990 (See Appendix), the eThekwini Municipal policy requires that
privately owned sites may be required to manage and make provision for their own
stormwater runoff.
The general level of management required is that of controlling all runoff emanating from
such a site in excess of that which would have occurred if the site was in its natural or
original virgin state. Such stormwater management may well be the provision of soak pits,
structures such as attenuation ponds or tanks (with controlled outlets where necessary, all to
ensure that the rate of runoff is reduced to predevelopment states and that runoff is not
concentrated onto adjacent neighbouring sites or other infrastructure, be it road drainage or
valley lines, streams etc.
1.1 Provision of soakpits
In general, where a site was originally developed with attenuation structures/ponds or

soakpits controlling runoff, these systems should remain. These sites are also
required to maintain and upgrade their systems to accommodate both existing and
any new additional hardening on the site.
Any application for approval of development on any open site without access to
municipal stormwater will be required to provide a stormwater management plan
based on a rational design method demonstrating that the soakpit/attenuation and
other proposed controls controlling the additional stormwater runoff generated due to
all the development of the site will not adversely affect downstream and neighbouring
areas. All designs are subject to the approval of the Coastal, Stormwater &
Catchment Management Department of the eThekwini Engineering Unit.

Additionally, existing single erven or properties that have a municipal stormwater
connection, but where the site has been developed to the point where more than 40%
of the site area has been hardened may be required to manage the excess runoff
generated on the site/erf itself.
Hardening of the site may be by either roofed areas or by otherwise generating more
stormwater runoff by providing paved driveways or other semi or impermeable
structures cumulatively covering in excess of 40 % of the area of the site or erf. If the
geological and stability profile of the site is suitable soakpits sized on the basis of the
excess hardened coverage of 40 % are to be provided or other engineered
attenuation measures based on an engineers design and the Rational method of
determining pre to post development runoff.
Many existing/old soakpits are very rudimentary e.g. sheet iron covered rubble filled
holes. These older soakpits had a effective lifespan rarely in excess of 10 years.
They cannot be cleaned or maintained and the initial storage volume provided was
generally minimal. Properly designed structures provide access to allow for essential
periodic removal of accumulated silt, organic material and other windblown materials
from the soakpit to allow it to continue to function effectively. Regular maintenance
extends the soakpits life span and maintains storage volume.
When drawings are submitted for approval of building additions and alterations and
the property had existing soakpits, then the developer must be able to demonstrate
that the existing soakpits are functioning. These existing soakpits will have over time
been filled with detritus and in many cases have been displaced by excavations for
pools or other structures etc. The original soakpits cannot be accepted unless they
were constructed properly and can be accessed for cleaning or inspection. Current
policy dictates that approval for any new addition on sites with soakpits must make
provision for installing new soakpits unless a professional engineer is able to inspect
the old/existing soakpits and certify the capacity and condition.
As such, new properly designed soakpits for stormwater shall be built, sized on the
basis of 1 (one) cubic metre of clear volume to drain each and every 40 square
metres of all roofed, paved or otherwise hardened areas on that site.
There are some alternatives to using soakpits and these include the provision of
rainwater tanks which can be used for watering gardens and/or engineered ponds.
These may be an attractive alternative to soakpits providing a stored source of
irrigation water. The acceptable storage ratio in terms of using rainwater tanks as an
alternative to soakpits is 60% of the volume of the tank. In other words, when
calculating the volume of storage required (on the 1 m3 to 40 m2 area) then 60% of the
rainwater tank volume may be claimed on the assumption that the tank is 40% full at
any given time.
Notes:
Hardened areas shall include roofed areas, all paving, surfaced driveways, pool
surrounds, etc.
Grass blocks or other permeable paving are considered as hardened areas but the
area can be reduced by up to 50% in such cases for determining the storage required
(i.e. as being actually hardened).
A typical detail of the soakpit design showing a plan and sectional view is to be
provided as part of any drawing submission for approval where a soakpit is to be
installed.

Drawings must table the extent of the hardened areas to demonstrate that the
required soakpit volume/s provided is adequate (i.e. 1 cubic metre of volume for every
40 square metres of area hardened).
Soakpits should be constructed 3 4 metres (minimum) from any structure or

boundary to prevent damage to buildings, foundations and to reduce a risk of piping
under boundary walls etc. Any reduction in this distance must be supported by a
geotechnical report.
All soakpits provided shall be drawn to scale in position on the plan view, labeled as
"new" together with the volume and overall dimensions.
The pipe drainage must collect and drain the rain water from collection points, (eg.
down pipes) to the soakpits and sized to ensure that the soakpits will fill during
significant storm events.
1.2 Provision of attenuation and other storage structures
Where economics, site geology, impermeable rock layers or clay soil types, slope
stability or other engineering considerations argue against the suitability of using a
soakpit then other attenuation controls or structures are mandated. On individual
erven which are suitable for soakpits, it may not generally require a professional civil
engineers design for soakpits for stormwater management.
Generally, for multi-unit residential or commercial developments or where coverage

exceeds 40% of an individual residential site, notwithstanding the suitability of the site
geology for soakpits, a stormwater management plan is required. This is based on the
increased risk of failure and higher percentage runoffs. Soakpits tend to fill in higher
order storm events before the peak stormwater runoff from the site is reached. In
other words the higher order peak runoff from the site is not reduced and the risk of
downstream flood damage is not diminished.
A competent professional civil engineer will be required to produce a Storm-water

Management plan modeling the run-off for the Pre and Post development scenarios
for at least the 1:10 and 1:50 year Recurrence Interval storms.
The Storm-water Management Plan provided must provide logical and

coherent detailed information allowing for verification of the engineers design.
The SMP must provide as a minimum, the information as listed in paragraph
1.3 below.
The pre-development runoff coefficient (C) should ideally be derived on the

component method where the slope component (Cs), soil permeability type
component (Cp) and vegetation cover component (Cv), are assessed and summed to
derive the runoff coefficient used for the site/drainage area as a whole.
The post development runoff coefficient (C) for the overall site can generally simply
be derived assessed on a proportional basis of the pre-development runoff coefficient
C and the post dev areas that have actually been hardened with C = 0.95. Some
examples are shown later in these notes.
Generally, a Time of concentration Tc of not less than 15 minutes for residential sites
and Tc of not less than 10 minutes for industrial sites must be used ( i.e. where the
calculated Tc is less).
The Rainfall statistical data to be used is available on the municipal website:

www.durban.gov.za/eThekwini/Services/Engineering/CSCM
This data is listed based on a latitude and longitude grid for a variety of recurrence
intervals from 2 to 1000 years for storm durations from 5 minutes to several days. The
data is given as point data (i.e. In mm of rain falling in the storm time period shown)
and must be converted to the rainfall intensity (mm/hr) corresponding to the Rational
formula.
Storage or attenuation/infiltration measures must provide for the difference between

the Pre development and Post development 50 year storm runoffs
generated/calculated.
The attenuation storage required can be assessed on a simple plot of the peak runoff
values Q calculated for the pre and post scenarios versus the Time/s of concentration
(Tc) for each scenario. i.e. plotting superimposed pre and post hydrographs as
simplified triangles.
For small catchments runoff is assumed to be zero at both Time = zero and at 2 x Tc
minutes. The volume of attenuation storage is represented by the area of the post
dev hydrograph which lies outside/above the pre dev hydrograph plotted. In effect any
runoff greater than that of the pre dev plot (50 year RI storm) must be
attenuated/retained.
The rate of outflow for any recurrence interval storm must be restricted to the pre-
development runoff. For example: runoff/outflow for the Post-developed site in a 10
year RI storm is not to exceed the 10 year pre-development runoff peak. The static
water head generated in the attenuation tank when the 10 year post development
calculated storage volume is reached should not result in a greater discharge from the
attenuation tank than the 10 year Pre-development runoff calculated. Similarly, the 50
year storm storage/outflow relationships must be attained.
The stormwater Management Plan must demonstrate with sufficient detail that the
proposed measures/structures/ponds can be provided and are practical and workable.
Where sites are steep the extended cut and fill banks which may result in providing a
pond or structure should be shown. Cut and fill banks obviously cannot extend into
building platforms or neighbouring properties and the drawings should demonstrate
that this has been catered for in the design.
The drainage system provided must be capable of delivering the volumes associated
with the attenuation structure. There is no point in providing for an attenuation pond or
other structure when the site layout and stormwater system is incapable of channeling
or conveying the increased and excess runoff generated. The increased runoff may
not bypass the attenuation provided and flow offsite/downstream without being
reduced to predevelopment rates of discharge.
Depending on the individual site characteristics and the location of the attenuation
feature, a degree of over-design may be inherently required to satisfy the requirement
of limiting runoff to pre-development conditions in the larger recurrence interval storm
events. Competent design can resolve such problems and minimise costs.
The use of level parking or flat areas as shallow attenuation tanks, creating wet or dry
lined or unlined ponds are all alternatives to costly conventional retention structures.
The objective is to maximize groundwater infiltration and/or to reduce peak runoff

from artificially hardened development further eroding streams and further
exacerbating the potential for flooding lower down in any catchment.

All Final stage/Construction drawings or building plans must include complete details
of all storm-water structures, all outlet controls, reticulation layout and pipe sizes and
all erosion protection measures required for construction. The architects submission
must incorporate the approved engineered stormwater management layout.
1.3 List of minimum information required in a stormwater management plan (for

both the pre and post development scenarios)
Street address and description of erf

Latitude and longitude or X, Y co-ordinates of site (Especially if no street address
exists)
Area of site
Length of estimated drainage flow path
Fall or slope of site
Runoff Coefficients used (Cv - vegetal cover, Cs - slope, Cp - permeability/soil
type)
Time of Concentration
Existing and future areas that have or will be developed (hardened)
Actual Runoff coefficients used eg:
o Roofs, premix driveways (C=0.95),
o paving (C=0.95),
o grass blocks (C=0.5)
Rainfall intensities I10, I50
Runoff calculations Q10, Q50
Attenuation volumes
All of the above should be tabulated for both the Pre and Post development for at
least the 10 and 50 year recurrence interval storms.

2 THE RATIONAL METHOD
2.1 Restriction on Usage
The Rational Method is still probably the quickest and most commonly used method
of estimating the peak runoff value of stormwater run-off generated from urban and
rural areas in spite of its limitations in application and accuracy. Municipal
guidelines/policy based on the National Building Regulations presupposes at least a
Rational method of determination.
The formula used in this method is
Q = ft x C x I x A/360 cumecs
Where
Q = the maximum/peak rate of run-off in cumecs (m3/s)

ft = an adjustment factor for the recurrence interval storm considered.
C= run-off coefficient (see applicable tables for determination)
I = the rainfall intensity (mm/hr)
A = area of catchment in hectares (1 ha = 10 000m2)
2.2 Area of catchment (A)
The area of catchment is the total area above a point of interest that will contribute to
the run-off at that point, either from naturally occurring stream flow or from overland
flow.
Such an area can be measured on a topographical map by drawing a line from the
outfall point and following the crests of ridges, spurs or high areas which can be
identified from contours and encloses those valleys or low areas which will drain
towards the outfall point. In manual mapping of any catchment, conceptually, runoff
flows as a vector line equally bisecting the angle formed when drawn through the
contour lines. This concept determines the extent of natural catchments.
The area of the catchment once defined on a topographic map is determined by

dividing the area into measurable geometric shapes (triangulation etc.), or by using
the measuring tools on the GIS.
2.3 Time of concentration (Tc)
The time of concentration can be regarded as the time it takes for the excess rainfall
resulting in run-off from the furthest significant part of a natural catchment to reach
the point being considered. The shape of the catchment has significant implications
on assessing the length of the flow path. For more accurate estimates of Tc it is
sometimes necessary to create several smaller catchments especially where the
shape of the catchment includes long narrow upper reaches which will not
significantly contribute to the overall volumes or peak flow generated.
It is difficult to accurately assess or determine the time of concentration with great

confidence since an iterative approach would be required to assess the depth of
sheet or concentrated flows with innumerable other considerations. However a
number of tables and equations are included for estimation purposes. To obtain a
measure of consistency, a general approach is recommended as follows:
(a) Sheet flow naturally concentrates becoming stream flow with greater flow depth
and generally higher flow velocities. The catchments size and shape is usually
significant in terms of determining the overland flow path to assess the time of
concentration. The flow path length must be assessed appropriately. The runoff
hydrograph for a long thin catchment differs significantly from a square catchment
with the same area and characteristics. The rates of runoff will differ and be
spread over longer periods of time. The total volume of runoff will tend to differ
also depending on the time surface water has to infiltrate into the ground. The
initial rates of infiltration depends on soil, moisture content and innumerable other
considerations.
(b) In assessing any flow-path when confronted with irregular shapes use your
engineering judgement to apportion a flow length that is appropriate, i.e. that
conforms to the runoff for the bulk of the area of the site to the point being
assessed. Similarly when assessing the slope of a catchment use a slope
conforming to the bulk or major portion of the site and use this in the Kerby
formula (see later). (NB. the slope component of the runoff coefficient is not what
is being considered here).
Guidelines suggest the use of the height difference on the 1085 method or equal area
methods. The 1085 height is the elevation difference measured from downstream at
the 10 and 85% points along the full length of the flow-path.
Whatever method is used it must be representative of the bulk of the catchment area
considered.
Since the methods used are subjective and inherently provide an estimate only it is
not necessary to exhaustively analyze most catchments in a municipal environment.
In large complex catchments the best method though laborious, is to break the
catchment down into smaller areas and individually assess and cumulatively add each
component areas runoff to derive a better estimate of the peak runoff of the whole. In
the municipal environment with generally small catchments or developments being
assessed, it is not necessary to exhaustively analyze but a sensitivity check should be
done.
Generally sheetflow only occurs in the first or initial say 200m portion of the upper end
of the catchment. Thereafter, a further time of flow in conduits/streams/gutters etc
becomes applicable and then Manning type equations or charts providing typical
velocities for various terrain types or lined conduits can be used to estimate the
balance of the travel time to the point under consideration.
There may be several different catchments contributing flow to the point under
consideration. The longest travel time calculated of all the different contributing runoff
routes to that point is assumed to be the time of concentration.
Similarly the area A used in the formula Q=CIA to determine the peak flow at that
point is the sum of the areas of all the catchments contributing runoff to that point.
Since the bulk of development in the city/metro region takes place where catchments
are relatively small (< 1 km square) we advocate using a minimum time of
concentration of 15 minutes for all undeveloped/rural/residential type sites. In other
words if the calculated time of concentration for a residential site is less than 15
minutes use 15 minutes, and where a site is predominantly hardened i.e. fully
developed commercial/industrial sites) then use a minimum of 10 mins if the
calculated time of concentration is less.
2.4 Run-off coefficient (C)

The run-off coefficient is a factor ranging between 0 and 1 which compensates for
variations in rainfall over the catchment, infiltration and overland flow velocity during a
storm, the shape of the catchment, ground slope, etc. Because of various
indeterminate factors including ground moisture content, vegetation, permeability of
soils, varying slopes, rainfall intensity etcetera, the coefficient 'C' is difficult to assess
and a widely diverging range of estimated runoff coefficients can result. To minimise
widely disparate results and allow for uniformity and consistency in approach, the
table method used by the Dept of Water Affairs & Forestry (DWAF) is to be used and
an excell version is available (shown below).
C can be derived from applicable tables for determination. Where several catchments
or sub-catchments contribute runoff to the point under consideration then unless C is
uniform for all (i.e. the same slopes and vegetation etc exist) then a modified C
applies which must be calculated as follows:
Coverall = (Sum of Ci x Ai)/(Sum of Ai) for all of the differing sub-catchments i.
For undeveloped sites the value for C must be derived from the sum of the
contributions of the ground slope Cs, the vegetative cover Cv and the permeability or
soil type Cp. In urban/industrial areas a combination of the percentage area
contribution of the hardened areas and the balance of the site area assessed in terms
of Cs, Cy and Cp above is logical/appropriate:
Ci for the catchment i = Csi + Cvi + Cpi
Ci may be considered to remain constant during any particular storm for smaller
catchments (<5 km2). However we do advocate using a modification factor ft to
reduce the runoff for lower order storms.
RI Storm year Reduction

Factor ft
2 year 0.5
5 year 0.55
10 year 0.6
20 year 0.67
50 year 0.83
100 year 1
But in a Pre and Post development assessment of runoff to estimate the attenuation
storage required to reduce post runoff, the difference in C between pre and post
development values of C determines the volume to be attenuated. The allowable
discharge should be determined by applying the above reduction factor in order to
obtain a lower value for the predevelopment peak flow since the rational peak flow is
generally considered to be conservative (high) for design of systems.
Values of 'C' for different types of catchment conditions and surfaces are shown
below. A spreadsheet calculator for estimating the C value is included in the later
examples in these guidelines.

CALCULATION OF RUNOFF COEFFICIENT DWA METHOD
PRE/RURAL Runoff Coefficient POST/URBAN Runoff Coefficient
RURAL URBAN %
Steepness/Slope Cs % > 900mm Lawn sandy<2% 0 0.08
< 3% 20 0.05 Lawn sandy>7% 0 0.18
3-10 % 50 0.11 Lawn heavy<2% 0 0.15
10 - 30 % 15 0.20 Lawn heavy>7% 0 0.30
> 30 % 15 0.30 Residential single 0 0.40
Cs 100 0.14 Flats/dense townships 0 0.60
Permeability Cp % Industry , light 0 0.65
Very perm (Dunes) 0 0.05 Industry , heavy 0 0.70
Perm (light soil) 10 0.10 Business local 0 0.60
Semi (most soils) 80 0.20 Business CBD 0 0.85
Imperm (rock, paving) 10 0.30 Streets/roofs 100 0.95
Cp 100 0.20 100 0.95
Vegetal growth Cv %
Dense bush, forest 10 0.05 AREA WEIGHTING FACTORS
Cult land, sparse bush 5 0.15 % DWA
Grassland 75 0.25 RURAL 0 0.57
Bare Surface 10 0.30 URBAN 100 0.95
Cv 100 0.23 LAKES 0 0.00
Cpost/desig
Ct = Cs + Cp + Cv = 0.57 n 100 0.95
In determining the value of 'C', the effect of both the present and future land use on
run-off must be considered if upgrading or new stormwater infrastructure is being
considered.
The Mean Annual Precipitation (MAP) for Durban/eThekwini region is predominantly

>= 900 mm MAP and even if the MAP is lower, we advocate using 900 MAP as a
minimum.
2.5 Rainfall Intensity 'I'
The rainfall intensity is the average rainfall in mm/hr for a design storm of a given
frequency having a duration equal to the Time of Concentration Tc. Rainfall figures
can be obtained from the eThekwini website once you have identified the appropriate
Latitude and Longitude for the site (available on the GIS). The figures given for the
times shown are point rainfalls (i.e. in mm) and must be converted to an intensity (i.e.
mm/hr)
www.durban.gov.za/eThekwini/Services/Engineering/CSCM.
Rainfall intensity figures are also included in the accompanying excell file.
In the design of a drainage system using the rational method, the duration of the
design storm is assumed to be equal to the time of concentration 'Tc' estimated at the
appropriate nodal point in the drainage system being designed.
'Tc' must be computed independently at each inlet or low spot. Thus, as the design
progresses downstream, the drainage area increases in size, 'Tc' will increase and as
a result, the rainfall intensity 'I' will correspondingly decrease.
Remember that the minimum value of Tc to be used is at least 15 mins (in rural or
urban residential areas) and 10 mins (in commercial/Industrial areas since the
hardened areas are proportionally much greater). In other words, where the actual
assessed/calculated Tc is less than 15 or 10 minutes then use 15 or 10 minutes as
appropriate as a minimum.
An excell file is also available giving rainfall intensities for storms of up to 8 hrs where
the latitude and longitude are known.
3 DESIGN OF A MINOR SYSTEM
3.1 Critical Points
The effectiveness of stormwater design depends largely on the identification in a

catchment area of those areas or points where flooding cannot be tolerated more than
once in 10 years due to the likelihood of heavy economic losses or social
inconvenience. Such points are termed critical points, and while they can sometimes
be pinpointed on a topocadastral map, they should always be identified during a field
inspection.
Critical points may occur:

(a) at low points in a road (where ponding will occur) and ponding water may overflow
the verge on the fill side, thus eroding fill embankments and flooding low-lying
property;
(b) at the intersection of a steep road with a flat road where water flowing down the
steep road could flood the intersection or overshoot the opposite verge;
(c) at the site of an important drainage structure eg. The confluence of a road with a
major stream. Where potential flooding of a development may cause high
economic losses.
Any design or planning for the overall system must entail an assessment of what if
scenarios. For example: ensure that you allow for an overland flood route below low
points in a system considering that the pipe system could fail. This may be allowing
for an access way or a footpath below such points.
3.2 Rainfall Intensity at Critical Points
At critical points, both sufficient inlet capacity and pipe capacity must be provided to
cope with the 10 year storm event. At all other points in a catchment, except for
certain cases discussed below, design generally is only for the 3 year storm.
In special cases e.g. in areas where uncontained stormwater resulting from the use of
a 3 year design storm would cause severe wash-a-ways in soft ground, the
Stormwater/Catchment Manager may require that the design be based on a storm
frequency of 5 years for non-critical points, and of 20 years for critical points.
In flat areas of the City with considerable commercial and industrial development, run-
offs should be based on a 10 year storm period. "Major" disposal systems (i.e. those
taking flows of about 10 cumecs or more) should be designed for 20 year storms and
in some special cases e.g. systems crossing Transnet/S.A.T.S. reserves, design
intensities may be required to be based on 50 or even 100 year storms.
3.3 Minimum Diameter
Note that downstream pipes should never be smaller in diameter than the upstream
pipe notwithstanding that hydraulic considerations (such as steeper hydraulic
gradelines or slopes) may support/allow this. Downstream pipes will obviously tend to
be blocked by any debris/objects transported down larger upstream pipes. The
minimum diameter of pipe shall be as follows:
300 mm in a servitude; and
375 mm in a road reserve.
3.4 Minimum Velocity and Gradient
The desirable minimum full flow velocity shall be 1,5 m/sec and the absolute minimum
full flow velocity should be 0,9 m/sec which is acceptable only in unusual
circumstances.
Desirable and absolute minimum gradients are shown in the following table:
Diameter Desirable Gradient 1/

300 80
375 110
450 140
525 170
600 200
675 240
750 280
825 320
900 350
1050 440
1200 520
3.5 Materials
In general, stormwater pipes shall have rubber ring joints, be spigot and socket spun
concrete pipes complying with S.A.B.S. 677 but fibre reinforced cement pipes are
permissible provided they comply with S.A.B.S. 819. Ogee type pipes are NOT
acceptable.
Other acceptable/suitable pipe types are : Weholite and Ribloc type pipes for use
where steep grades or to maximise the use of labour but their use is NOT
recommended in road reserves and road crossings .
3.6 Anchor Blocks
20 mPa concrete anchor blocks to details shown in Appendix 5 should be provided as

follows :
Grade (%) Spacing for 2,44 m pipe lengths

Over 50 every joint
30 to 50 inc. alternate joints
20 every 4th joint
10 every 8th joint
Spacing for intermediate grades can be interpolated.
3.7 Curved Alignment
In normal circumstances straight alignment between manholes should be used, but

curved horizontal alignment is acceptable subject to the following limitations:
(a) the minimum radius of curvature for an effective pipe length of 2,44 m is as
follows :
Pipe diameter (mm) Radius of curvature (m)

300 70
375 70
450 93
525 93
600 112
675 to 900 inc. 140
1 050 186
1 200 186
1 350 278
1 500 278
(b) curved alignment is only permissible with pipes having approved flexible joints.
3.8 Servitudes
The width of sewer and drainage servitudes is dependant upon the diameters of pipes
to be laid within the servitude area and should not normally be less than 2m.
However this width may be reduced at the discretion of the Deputy Head.
3.9 Position of Stormwater Sewers in Servitudes
Stormwater sewers in servitudes should be positioned as follows:

in 3 m servitudes - 1,0 m from a property boundary;
in 2 m servitudes - in the centre of the servitude.
3.10 Position of Stormwater Sewers in Road Reserves
Recommended layout of services in road reserves are shown in the Appendices for
various road widths. In existing roads already containing services, a stormwater
sewer should be laid in the verge at least 1 m clear in a horizontal direction from the
water main.
3.11 Manholes
Manholes should be placed at every change in horizontal and/or vertical direction or

at a maximum spacing of:
100 m for pipes up to and including 900 mm diameter;

150 m for pipes over 900 mm up to and including 1 200 mm diameter;
200 m for pipes over 1 200 mm in diameter.
Details of standard precast concrete ring manholes and brick manholes are shown on
drawing nos. 38570/1/2/3 and 38850/1/4.
Manholes may be constructed using 1 000 mm internal diameter class A precast

concrete ring units in accordance with SABS. 1294 to a maximum depth of 5 m for
pipes up to 375 mm diameter where a junction occurs and up to 600 mm with no
junction.
In all other cases including changes of direction and for manholes deeper than 5 m, a
specialised design is usually necessary. Provision is to be made for a landing in all
manholes deeper than 5 m in compliance with the requirements of the relevant
regulations including the Occupational Health and Safety requirements. Step irons are
considered unnecessary and should not be provided.
3.12 Manhole Covers
Where manholes occur in roadways, standard D.C. heavy duty cast iron covers and
frames in accordance with SABS. 558 Type 2B as shown on drawing number
DMW1281 should be used. Heavy duty precast concrete covers should be provided
in footways and verges and wherever vehicular traffic may be expected other than in
roadways. In all other cases light duty precast concrete covers are acceptable.
A new composite polymer specification manhole cover and frame has been field
tested and approved for use. Contact the CS&CM catchment or stormwater managers
for details.
Details of heavy duty and light duty precast concrete covers are shown on drawing
nos. 38853 and 38852 respectively.
Where manhole covers are to be sloped to suit road gradients, they should be laid on
shaped brickwork or in-situ concrete.
3.13 Benching in Manholes
All manholes should be benched with a smooth concrete channel formed to the soffit
of the pipe and every attempt should be made to streamline the "inlet to outlet" flow of
water.
3.14 Channels
The minimum roadway cross fall on any black top surface should be 2,5% and the
minimum longitudinal gradient should be 0,5% for concrete channels and 1% for
asphalt channels.
3.15 Minimum Cover
The minimum allowable depth of cover to the outside of the barrel of the pipe for
stormwater sewers is as follows:
(a) in servitudes 0,8 m

(b) in footways and verges 1,0 m below final kerb level
(c) in roadways 1,2 m below final constructed road level
If the required depth of cover cannot be achieved by importing additional material and
a lesser depth of cover is unavoidable, the pipe should be protected from damage, at
the discretion of the Deputy Head, by means of:
the placement of case-in-situ or precast concrete slab(s) over the pipe, isolated from
the pipe crown by a soil cushion of 100 mm minimum thickness. The protecting slab
(s) should be wide enough and so designed to prevent excessive superimposed loads
being transferred directly to the pipe (see Appendix);
the use of structurally stronger pipes able to withstand superimposed loads at the
depth concerned;
Where the depth of cover in roadways or footways and verges is less than 600 mm or
where the depth of cover in servitudes is less than 300 mm, protection of the pipe
from damage must be provided.
3.16Bedding
Bedding shall generally be in accordance with the requirements of Standard

Engineering Specification Part "DB" : Earthworks for Pipe Trenches.

The various bedding classes and their relevant load factors for rigid pipes are shown
in the appendix.
In saturated ground, steps should be taken to provide adequate drainage in trenches

with a minimum layer of 150 mm of 19 mm stone placed under the pipe. Should site
conditions warrant, filter fabric may be placed between in-situ material at the trench
bottom and the stone mat. A cut-off drain placed at the seepage side of the trench
bottom and connected to a stormwater manhole may be necessary. Further
information on bedding of pipes is contained in the "Concrete Pipe Handbook"
published by the Concrete Society of Southern Africa, a copy of which is available in
the Technical Library of the Engineering Unit.
3.17 Invert Levels at Manholes
General
The standard practice of laying pipes "crown to crown" with benching formed to half-
pipe level normally creates no problems but in some cases insufficient attention is
given to the adjustment of invert levels at manholes to compensate for loss of energy
between inlet and outlet pipes to prevent surcharge occurring in the manhole. In other
words the water level rises in the manhole to compensate for the loss of energy and in
extreme cases, the water level rise may be more than the manhole height and water
surges out of the manhole dislodging the manhole lid.
Standard Conditions
At manholes with no drop inlets or no substantial side inflows or where the change in
horizontal direction is less than 4 the difference in invert levels is determined by the
difference in pipe diameters only provided pipes are laid "crown to crown" and
benching complies with the requirements of paragraph 3.13 above.
Special Conditions
In the following circumstances conditions within a manhole warrant detailed

examination:
(a) where the velocity head from the inlet pipe is destroyed e.g. at a drop manhole;
(b) where a relatively large inflow enters a manhole from an inlet or from one or
more subsidiary lines.
When considering (a) and (b) above, the following criteria should be taken into
account in calculating the required invert level of the outlet pipe:
(i) full pipe flow at entry to outlet pipe;
(ii) the water level in the manhole is not to be above crown level of the pipe
carrying the major incoming flow and the crown of other incoming pipes should
not be lower than this level.
The velocity head required at the entry to the outlet pipe above the crown (the top or
soffit) of the pipe =V/2g
where V = required velocity at entry to outlet pipe in m/s, and g = 9,81 m/s, and in
order to satisfy conditions (i) and (ii), the difference in invert level, between the inlet
and outlet pipes = V/2g + difference in diameter of pipes.
Note: the downstream pipe should not be a smaller diameter than the upstream pipe
notwithstanding that the available grade may allow for this. This is to prevent
potential blockages where an obstruction enters the system upstream).
Bend Losses
Bend losses should be taken into account where there is a change in horizontal
direction greater than 4 and although opinions vary on the extent of such losses in
manholes, a loss of 50% of the velocity head of the inlet pipe is considered
reasonable.
For a 90 bend
Velocity head at inlet pipe = Vi/2g

Bend loss = 0.5 x Vi/2g
Velocity head at outlet pipe = Vo/2g
Available velocity head downstream of the bend = Vi/2g - 0.5 Vi/2g= 0.5 x Vi/2g
The difference in invert level between inlet and outlet pipes equals the difference in
pipe diameters plus the difference between required and available velocity head.
=difference in pipe diameters + Vo/2g - 0.5 x Vi/2g
For a 45 bend:
From the graph in Appendix 12, the available velocity head downstream of the bend
=Vi/2g - 0.75 x 0.5 Vi/2g = 5/8 x Vi/2g
and the required difference in invert level
= difference in pipe diameters + Vo/2g -5/8 x Vi/2g
Where a substantial inflow from a subsidiary line occurs at a bend, the conditions
detailed in clause 3.17.3 apply.
3.18 Inlets
Where the grade of a road is flatter than 0,5%, graded channels are used and single
inlets are provided at 30 m intervals. In this Service Unit, for roads having grades
steeper than 1 in 200, the usual practice is to provide inlets at 40 m with depressed
channels extending 1 m on the upstream approach.
A more economical design could be obtained if more detailed investigations were to

be conducted into inlet opening length, upstream channel length, road grade and the
width of stream flow in the road and to encourage such analysis, Charts have been
included in the appendix to these guidelines.
These charts are based on the following:
(a) maximum allowable stream width to be 2,5 m and at critical points, 3,2 m;
(b) interception at inlets to be 80% of flow and at critical points, 100% of flow
(c) depressed channels to be provided regardless of the type of kerb used; and
(d) inlets to have openings in units of 1,2 m varying in length up to 5,6 m
(N.B. A 2 bay inlet consists of 2 No. 1.2 metre wide units but after deducting the
thickness of brickwork i.e. 0,2 m, the waterway area is reduced to 2 m).
4 SELECTION OF CLASS OF PIPE

The class of pipe to be used can be obtained from the "Concrete Pipe and Portal
Culvert Handbook" issued by the Concrete Manufacturers Association.
5 PARTIAL FLOW IN PIPES
In the design of a new reticulation system, it is assumed that pipes are flowing full and
partial flow is not normally considered. However, it is sometimes necessary to know
the velocity and discharge of pipes flowing partially full and in such cases, the graph
in Appendix 19 may be used or the Manning Equation which is included in the
accompanying spreadsheet.
6 SUBSOIL DRAINS
A subsoil drain should consist of either of the following:
(a) for small volumes of seepage water, 19 mm or 25 mm grade single size stone as
per SABS 1083 wrapped in drainage grade filter fabric with a 200 mm overlap at
the top of the drain to form a nominal 200 mm by 200 mm square section with a
100 mm layer of coarse clean sand placed on either side of the 200 mm x 200 mm
section;
(b) for regular and high seepage flows, subsoil pipes wrapped in drainage grade filter
fabric with a minimum overlap of 100 mm situated at the top of the pipe and
covered with a clean coarse clean sand compacted to 95% Mod AASHTO.
It is often convenient to incorporate a subsoil drainage system in the same trench as

the stormwater sewer, running alongside or in the road. This technique allows for
easy interception of transverse subsoil drains from under the road. If a subsoil drain
is piped using perforated pitch fibre, slotted concrete or no-fines concrete pipes, it can
be connected to a stormwater manhole or inlet but it is advisable to use fibre cement
or spun concrete pipes for the connection through the brickwork.
Where subsoil drainage is required to cut off seepage, e.g. under a road, it may be
connected into a conveniently situated stormwater manhole or catchpit by means of a
no-fines concrete block built into the side wall of the brick chamber instead of bricks.
7 LARGE STORAGE PONDS
A major drainage system may consist of natural and artificial watercourses, large
conduits, stormwater storage facilities, servitudes and floodplains. Should such a
system be intended to cope with storms of 100-year frequency, the severity of the
storm and consequent disruption of certain activities may allow playing fields,
carparks, open spaces and similar areas to be used for on-site storage of stormwater.
Such storage ponds can be used to regulate the rate of run-off and also for the
control of pollution. They are classified as either attenuation (retarding) ponds, when
only a slowing down in the rate of run-off is required, or retention ponds, when, in the
absence of a positive outlet, the run-off is retained for future use.
A major drainage system incorporating storage ponds will only operate infrequently,
the more so in Durban because of the topography and consequently the design of
such systems is rarely necessary. However, if required, more information on the
design of storage ponds can be found in Part D, Urban Stormwater Management, of
the Guidelines for the Provision of Engineering Services for Residential Townships, a
copy of which is available in the Technical Library of the Engineering Unit.

8 DESIGN OF SHORT LENGTH CULVERTS UNDER EMBANKMENTS
The characteristics of flow in a culvert are complicated since the flow is controlled by
many variables, including inlet geometry, slope, size, roughness, approach, tailwater
conditions etc., and to obtain an accurate determination of flow would require
laboratory or field investigations.
An estimate of the capacity of a culvert CANNOT be made using the Manning (or
similar) formula for steady state pipe flow as the result would, in general, give a
culvert size which is far too small.
However, an approximate solution to the problem can be obtained using the following
charts and formula. However, in all cases, the design of culverts under embankments
should either be checked by, or referred to, the Coastal, Stormwater & Catchment
Management Department of the Engineering Unit.
9 CALCULATING TIME OF CONCENTRATION
Note: Where the calculated value of Tc is less, then use the following minimum values
for Tc
Thick vegetation (bush, wooded areas) 15 mins

Cultivated areas and parks 15 mins
Residential areas 15 mins
Fully developed, commercial, industrial 10 mins
Time of Overland Flow
The KERBY formula can be used where flow is overland/sheetflow, ie streamflow is

not yet developed:
Tc (in minutes) = 36x (r x L/1000/S0.5)0,467

Time of overland flow Units
Type of surface r factor Tc - Mins
Smooth paving 0.02 L - Distance (m) not greater than 200m
clean soil 0.10 s - Slope (m/m)
sparse grass 0.30 r - roughness value for type of surface
mod grass 0.40 EXAMPLE
thick bush/grass 0.80
Tc overland MINS 19.96
The above table is not to be used for an overland distance (L) where L is a lot greater
than 200 m. Here streamflow would be well established. Use stream flow equation
and/or the manning equation and/or table values below in combination with the Kerby
equation where L exceeds 200m.
Bransby Williams Streamflow equation -Natural Watercourses only

Tc (in minutes) = 60(.87L2/1000000000/S)0.385
This table gives a guide to the time taken with slope and distance shown
The units are the same as tabulated above and yield the following typical results
% slope Slope length (m)
m/m 100 200 400 500 800 1000
1 0.01 4.0 6.8 11.6 13.7 19.7 23.4
2 0.02 3.0 5.2 8.9 10.5 15.1 17.9
15 0.15 1.4 2.4 4.1 4.8 7.0 8.3
30 0.3 1.1 1.8 3.1 3.7 5.3 6.3
50 0.5 0.9 1.5 2.6 3.0 4.4 5.2
The following table can be used to estimate the lower limit of the flow time because
the table gives relatively high velocity values for large catchments indicative of
established runoff/flows.
grassed forest/me
sheet flow waterwa near bare adow
% slope paved (m/s) y (m/s) ground (m/s) lawn (m/s) cultivated (m/s) (m/s)
1 0.6 0.45 0.3 0.2 0.15 0.1
2 0.8 0.65 0.4 0.3 0.2 0.1
5 1.4 1 0.7 0.5 0.3 0.2
10 1.9 1.45 0.9 0.7 0.45 0.25
15 2.4 1.8 1.1 0.8 0.55 0.3
20 2.7 2.05 1.3 0.9 0.65 0.35
25 3 2.3 1.4 1 0.7 0.35
As examples:
the depth of sheet flow to achieve the paved velocities shown above would be approx
20 mm.
the depth of flow in a natural stream sloping at 1 % would be some 200 mm deep to
reach a velocity of 0.45m/s.
the depth of flow would be approx 50 mm in the case of the grassed waterway to
reach the velocities shown
10 EXAMPLES OF CALCULATING THE RUNOFF COEFFICIENT C
In a rural (undeveloped) area the catchment being assessed has slopes where 20%
of the area < 3%, 50% is between 3-10%, 15 % of the site area has slopes beween
10-30% and the remaining 15% of the area is very steep with a slope of> 30%.
Entering these values will generate a runoff slope component (Cs) of 0.14.
on the same site/catchment the soil profiles/permeability characteristics over the

surface show 0% of the area is very permeable, 10% of the area is permeable (light
soil), 80% of the site is semi-permeable (most soil types), and the remaining 10% of
the area is visible sheet rock. Entering these values will generate a runoff permeability
component (Cp) of 0.20.
on the same site/catchment the vegetation profiles across the surface shows 10% of
the area is dense bush, 5% of the area is cultivated land or sparse bush, 75% of the
site is grassland, and the remaining 10% of the area is bare rock. Entering these
values will generate a runoff vegetal cover component (Cv) of 0.23.
The cumulative runoff component for the entire site (C)=(Cs)+(Cp)+(Cv) =

0.14+0.2+0.23=0.57
(See the following table)

DWA METHOD
PRE/RURAL Runoff Coefficient POST/URBAN Runoff Coefficient
Catchment MAP MAP URBAN %
Catchment Slope CS % > 900mm Lawn sandy<2% 0 0.08
< 3% 20 0.05 Lawn sandy>7% 0 0.18
3-10 % 50 0.11 Lawn heavy<2% 0 0.15
10 - 30 % 15 0.20 Lawn heavy>7% 33 0.30
> 30 % 15 0.30 Residential single 0 0.50
100 0.14 Flats/dense townships 0 0.60
Soil Permeability Cp % Industry , light 0 0.65
Very perm (Dunes) 0 0.05 Industry , heavy 0 0.70
Perm (light soil) 10 0.10 Business local 0 0.60
Semi (most soils) 80 0.20 Business CBD 0 0.85
Imperm (rock, paving) 10 0.30 Streets/roofs 67 0.95
100 0.20 100 0.74
Vegetal growth Cv %
Dense bush, forest 10 0.05 AREA WEIGHTING FACTORS
Cult land, sparse bush 5 0.15 % DWA
Grassland 75 0.25 RURAL 40.00 0.57
Bare Surface 10 0.30 URBAN 60.00 0.74
100 0.23 LAKES 0 0.00
Rural Catchment coeffCt = 0.57 Cdesign 100 0.67
At a later stage the same site is developed as follows:
The site is partially developed with 40% of the area hardened (being roof, access
roads and paving), 20 % of the site is landscaped into flat lawns, and the balance
remains in a pre-development condition.
In the above excell table above, the POST/Urban portion of the table, 40% hardening
of the overall site equates to 67% of the developed/Post portion, and the 20%
landscaped portion equates to the balance or 33% of the developed/Post portion.
The post development runoff coefficient is a combination of 40% of the site with a
runoff C = 0.95, 20% of the site with C=0.3 say, and 40% of the site with say the
original undeveloped C=0.57 as previously calculated/shown. The combined C can
simply be estimated by the calculation C= (40% x 0.95 + 20% x 0.3 +40% x 0.57) /
100 = 0.67.
11 METHOD OF STORMWATER RETICULATION DESIGN
1 Identify the catchment area on a topographical plan and locate the natural
outfall.
2 Determine the points in the catchment at which flooding is intolerable and

mark these on the plan as critical points.
3 Build up a provisional network of drainage on the plan, starting at the top of

the catchment and working downwards until the outlet point is reached.
4 At this stage the field inspection should be done. After the network has been
checked and adjusted, if necessary, to suit local conditions, the stormwater
lines are surveyed in order to establish levels, lengths, grades, etc.
5 Mark on the plan the position of those inlets which are obviously necessary
(e.g. at local low points, intersections, critical points, etc.) And judge where
intermediate inlets are required. Rough calculations of local flows and
reference to the Kerb Inlet Charts : Appendices 13, 14 and 15 should be made
to assess spacings.
6 Establish the critical flow paths; i.e the flow paths from each critical point to the
outlet point of the catchment. These define the spinal system of critical
pipelines which is required to carry the 10 year storm flows from each critical
point to the outlet point.
7 Code the reticulation system so that each branch of the network and each
length of pipe between manholes or inlets of any branch is indentifiable : for
example A-B would represent the pipe length between manholes or inlets A-
B on Figure 1 of this Appendix.
8 Calculate the area of catchment above the inlet.
9 Determine the time of concentration Tc in minutes.
N.B. The longest time of flow for all of the catchments (upstream catchments) that
contribute runoff to the point considered must be used when calculating the design
runoff at that point. If two or more catchments contribute runoff to the point in
consideration, then the flow time for each catchment should be assessed. The
flowtime will generally consist of an overland flow time, any streamflow (in larger
catchments) and or any conduit flowtime (in pipes or channels etc). The longest time
estimated for all of the catchments draining to the point will be the Tc used to
determine the rainfall intensity to be used in the formula Q=CIA/360. The area used in
the calculation will be the cumulative areas of all of the upstream catchments which
drain to the point in consideration.
12 GENERAL METHODOLOGY
Label each catchment eg Area 1, Area 2 etc.
Determine the overland flow length and average slope for each catchment.
Determine the time of overland flow (use Kerby but L not > 200m) + time of stream
flow (for distance > 200 m if applicable) + time of flow in conduit (if applicable then
use chart or use Manning equation).
For subsequent downstream catchments the time of overland flow (length not >
200m) + time of stream flow (if applicable) + time of flow in conduit must be
determined when considering the design flow.
NB. If Tc < 15 mins (rural/parks/residential) use 15 mins
OR
If Tc < 10 mins (commercial/industrial/largely hardened sites) use 10 mins
Calculate the relevant runoff co-efficient C for each catchment and for the design of
pipes downstream, the area and longest time of concentration for all the contributing
catchments must be used.
Where more than one catchment contributes flow and the runoff Coefficient differs for
each contributing Catchment then a mean or representative C value must be
determined on a proportional area basis i.e. C = ((C1 x A1) + (Ci x Ai) etc.) / (A1 +Ai
etc)

Identify the Latitude and Longitude of the site. (This can be located using the GIS)
and obtain the relevant intensity of rainfall for the appropriate storm return period from
the municipal website or the accompanying rainfall intensity excell spreadsheet.
www.durban.gov.za/eThekwini/Services/Engineering/CSCM
Then calculate the PEAK runoff from Q = CIA/360 cumecs
Using the field information, pipe inverts between manholes/inlets are used to calculate
pipe gradients and a nomogram or manning formula can be used to establish the
following information:
theoretical pipe diameter;

selected pipe diameter;
capacity of selected pipe diameter; and
velocity of flow
All the above data should be tabulated for each step in the calculation but do not
proceed downstream of a junction until all the adjoining branches have been dealt
with.
When the first leg below a junction is considered in the calculation, the areas of each
catchment area for each branch leading to the junction are added together to give the
total area for this leg. This procedure of summing the areas ensures that critical lines
are carrying the 10 year runoffs from the full contributing catchment areas to the
critical point considered.
Using the flow value, you can derive the most suitable inlet from the Kerb Inlet Charts
given in the Appendix bearing in mind the following:
(a) unless the position of an inlet is fixed by physical conditions, the limiting factor
to be used is the maximum allowable stream width of 2,5 m (3,2 m at critical
points) which will often lead to greater spacing between inlets and, as a result
a more economical design but care should be taken that the intervening
spacing is carefully considered in areas where resulting flood damage could
occur.
(c) the bypass flow should normally be about 20% of the calculated runoff unless total
collection is desired (e.g. at a critical or local low point) and the appropriate inlet
for approximately 80% interception is selected from the relevant Kerb Inlet Chart.
The information should be tabulated on your sketch plan of the site for ease of later
reference and for review.
At low points select an inlet for a flooded width of 3,2 m (see examples on inlet
charts).
When necessary calculate energy losses at manholes using the information given in
Paragraph 3.17 above and tabulate the information on your design sketch plan layout.
It is important to note that the basic method of runoff calculations, pipe and inlet
design shown in this text, may be adopted for use in designing extensions to existing
drainage systems or upsizing of pipes in existing networks.
However, it is imperative that the design of extensions or relays should not overload
the capacity of existing downstream pipes. If this situation occurs, please consult the
Stormwater/Catchment Manager for assistance. Attenuation storage or upgrading
downstream may be required.

13 WORKED EXAMPLE

In the above layout, the low point is a critical point inlet C, the critical path is C-D-E-
F.
Consider point of entry at A
Area drained/Catchment Area is Area 1= 2 ha (= 20 000 m2)

Flow Length =150 m
Height of Fall = 4 m
Average Grade = 4/150 = 0.03 or 3 %
Tc (Kerby formula in minutes) = 36x (r x L/1000/S0.5)0,467
Tc (Using S=0.03, r = 0.4; L = 150) yields Tc = 22.5 minutes
For rural/residential areas Tc should be >= 15 mins therefore Tc = 22.5 mins is

in order.
Estimate Runoff Co-eff for Area 1 ie. C1 = Cs + Cp + Cr = 0.05 + 0.2 + 0.25 =

0.5
Design is normally for a 3 year storm.
NOTE:
In this example rainfall for a 5 year storm return period has been used
but the same principles otherwise apply for a 3 year storm with the
proviso that storm rainfall intensity would have to be interpolated
between the 2 and 5 year data since 3 year data has not been included.
The site is at Latitude 29 deg 44 min longitude 30 deg 50 min
To find the rainfall intensity we must interpolate for the Tc calculated of 22.5
mins between the values shown in the table for the above Latitude and
longitude between 15 min and 30 mins. The formula used is of the form ITc =
(Tc - 15)/(30 - 15)*(I30 - I15) + I15
Tc = 22.5 minutes gives I = 98.5 say 99 mm/hr
The recurrence interval adjustment factor ft for 5 year storm is 0.55 from the
table in paragraph 2.4 above.
Q = ft x CIA/360 = 0.55 x 0,5 x 99 x 2/360 = 0.151 cumecs
From information obtained in the field, the grade between A & B is 1 in 42.
The slope S is therefore = 0.024
From Manning equation Q = V x A = 1/n x (R)2/3 x S0.5 x A = 1/n x(D/4)2/3 x

S0.5 x Pi x D2/4
The manning n value for a concrete pipe is between 0.011 new to 0.013 for an
older pipe
The theoretical pipe diameter = 300 mm using Manning or find the diameter
using the pipe flow charts.
The capacity of the selected pipe (new and flowing full) = 0.180 cumecs
The headwall depth required for overland flow to enter the pipe can be found
in the appendix chart or estimated in the following spreadsheet calculator:
SHORT sloped PIPE CULVERTS <= 250 mm dia
HEAD H DIA D (m) H/D slope
1.25 0.09 13.88888889 0.001
q=D2(gD)0.50.48(S/0.4)0.05(H/D)1.9 m3/s H/D>0.8 0<h/d<0.8
q=D2(gD)0.50.44(S/0.4)0.05(H/D)1.5 m3/s 0.8>H/D>1.2 0.8<H/D<=1.2
q=0.6pi (D2/4)(2gh)0.5 m3/s 0.018902997 h/d>>1.2
In the above calculator, a head of 0.7 m only is required to induce a flow of

0.157 m3/s in a 300 dia pipe. The actual available head would normally be
greater in practice since the minimum cover for a pipe is more.
The velocity of flow in selected pipe at full pipe flow conditions =2.5 m/s
Before considering conditions at inlet B, calculate the time of flow in pipe A-B
For A-B Flow Length = 80 m
Flow Velocity in 300 mm dia Pipe at Grade of 1 in 42 = 2,5 m/s
Flow time = Length of pipe / velocity = 80/2.5 = 0.5 mins
The time of concentration Tc at B = 22.5 mins + 0.5 mins = 23 mins
Consider Inlet at B
Area drained A = 0,5 ha

Flow Length = 120 m
Average Grade = 5/120 = 0.041 = 4.1%
Tc (Using Kerby Formula S=0.041, R= 0.4; L = 120) yields Tc = 18.3 minutes

OK
Estimate Runoff Co-eff = C2 = Cs + Cp + Cr = 0.11 + 0.2 + 0.25 = 0.56
Assume 5 Year Storm Return Period and I = 110 mm/hr by interpolating

between the tabulated values for I at 15mins and 30 mins because Tc = 18.3
mins.
Q = ft x CIA/360 = 0.55 x 0.5 x 110 x 0.5 /360 =0.035 m3/s
With a grade of 1 in 20 between B1 and B and using manning formula, the

selected pipe diameter = 300mm
capacity of selected pipe = 0,392 cumecs
the velocity of flow in the selected pipe is negligible.
Consider MH.B

Area contributing to MH B = Area 1 + Area 2 = 2.5 ha
Calculate the longest travel time for contributing areas to MH B gives Tc = 23

mins
The combined runoff coeff C for Area 1 and 2 = (C1 x Area1 + C2 x Area2)/
(Area1 +Area2) since different runoff coefficients were determined for Area 1
and Area 2 (i.e we are calculating a mean representative value) = 0.51
Rainfall intensity I for a 23 min 5 year storm interpolated between the statistical
intensities given for 15 mins and at 30 mins results in I = 93.1 mm/hr
Q = ft xCxIxA/360= 0.55x0.51x93.1x2.5/360 = 0.181 cumecs (= 0 181 m3/s or

181 litres/sec)
The surveyed grade is 1 in 80 between B and C
selected pipe diameter = 375 mm
capacity of selected pipe from manning or table = 0,230 m3/s

the velocity of flow in the selected pipe = 2,1 m/s
Length of Pipe B-C = 60 m
Flow Time in Pipe B-C = 0.5 mins
Tc at C = 23.5 mins
Consider Inlet C
NB: C is a lowpoint i.e design is for a 10 yr storm as a critical point
Area drained = 1.7 ha

Flow Length = 180 m
Average Grade = 14/120 = 0.12 or 12 %
Tc (Using Kerby Formula S=0.12, R= 0.4; L = 180) yields Tc = 17.3 minutes

OK
Estimate Runoff Co-eff = C2 = Cs + Cp + Cr = 0.2 + 0.2 + 0.25 = 0.65
As the runoff entering inlet C from Area 3 will be required in the sizing of the
outlet, Q should now be calculated for this area.
A = 1.7 ha
Tc = 17.3 mins
C = 0.6 as calculated above
I for 10yr storm = 142.4 mm/hr (use 10 yr intensity being a critical point)
Q = ft 10year xCxIxA/360= 0.6 x 0.65 x 142.4 x 1.7/360 = 0.262 cumecs (= 0.262
m3/s or 262 litres/sec)
For total flow at C
Area Ac = Area1 + Area2 + Area3 = 4.2 ha

Tc = Tc(of Area 1) + Tc(from point A to B) +Tc(from point B to C)=
22.5+0.5+0.5=23.5mins (being longest Tc for all contributing areas)
Mean or representative runoff coeff C at point C must be determined. It is a

proportional average of all 3 of the contributing catchments, as follows:
= (C1xA1 + C2xA2 + C3xA3)/(A1+A2+A3) =
(0.5*2+0.56*0.5+0.65*1.7)/(2+0.5+1.7)
= 0.57
Intensity I10year = 120.8 mm/hr being the inter[polated value for 23.5mins for a 10
yr storm
Q = ft 10year x CABCx I xAABC/360= 0.6*0.57*120.8*4.2/360 = 0.482 cumecs (=

0.482 m3/s or 482 litres/sec)
With a grade of 1 in 75 between C and D
Selected pipe diameter from manning equation or pipe flowchart = 535 mm dia
pipe (Capacity full flow = 0.592 l/s)
capacity of selected pipe = 0.71 cumecs
velocity of flow in selected pipe = 2.5 m/s
Length of Pipe C-D = 50 m

Flow Time in Pipe = Length/Vel = 50/2.5 = 0.33 mins
Tc at D = Tc at C + 0.33 mins = 23.5 + 0.33 = 23.8 mins
Consider MH.D
Area = A1 + A2 + A3 as for inlet C = 4.2 ha
Runoff coeff = 0.57 as at pt C
Intensity I10year = 119.8 mm/hr from interpolation for Tc = 23.8 mins for 10 year
storm (C was a low pt or critical pt)
Q = ft10year xCABCD x I x AABCD/360

= 0.6*0.57*119.8*4.2/360
= 0.478 cumecs (= 0.478 m3/s or 478 litres/sec)
With a grade of 1 in 80 between D and E
selected pipe diameter = 600 mm

Capacity of selected pipe = 0.54 cumecs
velocity of flow = 1.91 m/s
Length of Pipe D-E = 100 m
Flow Time in Pipe = 100/1.9 = 0.9 mins
Tc at E = 23.8 + 0.9 = 24.7 mins
Consider Inlet E
Flow Length = 70 m
Average Grade =3/70 = 0.04 = 4 %
Tc (Using Kerby Formula S=0.04, R= 0.4; L = 70) yields Tc = 14.4 minutes say
15 mins minimum
Estimate Runoff Co-eff for Area 4

= Cs + Cp + Cr
= 0.11 + 0.2 + 0.25
= 0.56
For Total Flow at E
Area = A1 + A2 + A3 + A4 = 4.5 ha
Tc = 24.7 mins
C = (C1xA1+C2xA2+C3xA3+C4xA4)/(A1+A2+A3+A4)
=(0.5*2+0.56*0.5+0.65*1.7+0.56*0.3)/(2+0.5+1.7+0.3)
= 0.57
Intensity I10year = 116.6 mm/hr by interpolation for Tc = 24.7 mins
Q = ft10year xCABCD x I x AABCD/360

= 0.6*0.57*116.6*4.5/360
= 0.498 say 0.5 cumecs (= 0.5 m3/s or 500litres/sec)
With a grade of 1 in 10 between E and F

selected pipe diameter = 600 mm (pipe size must conform to upstream pipe
size even though a 375 mm pipe has the reqd capacity at a grade of 1/10)
capacity of selected pipe = 2.3 cumecs
velocity of flow = 8.1 m/s

14 DESIGN OF STORMWATER KERB INLETS
Once the pipe reticulation designed, consideration should be given to the sizing of
inlets. Examples are shown on the actual charts.
Briefly considering the above example comments are as follows:
Consider inlet B1
Q has been calculated above

The gradient of the Road must be known
Note that normally 80% of gutter flow is to be intercepted
The Maximum Stream Width = 2.5 m
The road Crossfall should be determined
Using kerb inlet design charts, the number of bays and upstream channel
length can be determined
Check that the Stream Width is not exceeded and determine the actual
intercepted flow (the balance should be added to the required flow at the next
downstream inlet Data should be tabulated or put on your design sketch layout
plan
Consider Inlet C
Q will be as calculated plus the bypass flow from inlet B1
As this inlet is at a low point, the approach grade on either side of the inlet will
be reasonably flat and a grade of 1% can be assumed in both cases.
Interception must be 100% and the maximum stream width should not exceed
3,2 m.
If flow is large, auxiliary inlets should be provided to intercept part of the

upstream flow.
Using the chart for low point conditions Note that at low points there is no
necessity to depress the upstream channel.
Consider Inlet E
Area contributing flow is area 4 only and the Tc for area 4 will be used in the
calculation only.
(N.B. a storm return period of 10 years is used only in the design of the pipe
network) due to the low point or critical point at C
There is no bypass flow from inlet C
Conditions at Point of Entry in Area 1
The available head H must be determined on site = 0,94 m and site conditions
are assumed to be such that this depth of headwater is acceptable

15 CALCULATIONS OF INVERT LEVELS AT MANHOLES
SEE PARAGRAPH 3.17 ON PAGE 17
In the previous example:
Manhole B
if there was a substantial flow from inlet B1 then at MH B it would be assumed that
the velocity head of the inlet pipe from A was destroyed and consequently bend
losses would in such a case be irrelevant and the reduced invert level would be
calculated as follows:
The required difference in invert level required is
= (dout - din) + vout2/2g
NOTES:
Vout is the velocity of flow in the outlet pipe using the calculated design runoff
flow divided by the full area of the chosen pipe
The velocity head would be measured above the crown of the pipe i.e.
assuming full pipe flow
If there were no inflows but only the energy losses due to the 900 bend of 0.5 x v2/2g.
Then the required difference in invert level required would be:
= (dout - din) + vout2/2g (vin2/2g - 0.5vin2/2g)
Inlet C11
At this inlet the change in grade of the pipe is small and can be disregarded but as
there is a substantial inflow, a energy or velocity head adjustment is required.
As there is a difference in pipe diameter, the reduced invert level of the outlet pipe
from the exit level of the inlet pipe, is the sum of the velocity head required for the
outlet pipe and the difference in diameter.
i.e The required difference in invert level required is

= (dout - din) + vout2/2g
Manhole D
At manhole D where there is a 900 bend and no difference in pipe diameter and no
inflow, the difference in invert level is simply = 0.5vin2/2g
Inlet E
At this point the effect of the small inflow can be disregarded and the required
difference in invert level required is
= (dout - din) + vout2/2g - (vin2/2g - 0.75 x 0.5vin2/2g)
Data should be tabulated in on a sketch of the layout or in a suitable table.

APPENDICES
PART R Stormwater Disposal SABS 0400 - 1990 page 36
Hydraulic elements of circular sections page 38
Bend losses page 39
Details of anchor blocks page 40
Discharge through box culverts : inlet control page 41
Discharge through pipe culverts : inlet control page 42
Kerb Inlet chart 1: cross fall 2% page 43
Kerb Inlet chart 2: cross fall 2,5% page 44
Kerb Inlet chart 3 & 4 : cross fall 4% and 6% page 45
Pipe bedding details page 46
Protection of pipes at reduced depths of cover page 47
Recommended layout of services - 9.5 metre reserve page
48


4 Ethekwini Design Manualmay 2008

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

4 Ethekwini Design Manualmay 2008

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

4 Ethekwini Design Manualmay 2008

Uploaded by

Copyright:

Available Formats

ENGINEERING UNIT

COASTAL STORMWATER & CATCHMENT MANAGEMENT DEPARTMENT

Revision Date: May 2008 1

DESIGN MANUAL: GUIDELINES AND POLICY FOR THE DESIGN OF

Revision Date: May 2008 2

1.1 Provision of soakpits

2 THE RATIONAL METHOD page 10

2.1 Restriction on usage

3 DESIGN OF A MINOR SYSTEM page 14

3.1 Critical points

4 SELECTION OF CLASS OF PIPE page 20

5 PARTIAL FLOW IN PIPES page 20

6 SUBSOIL DRAINS page 20

7 STORAGE PONDS page 20

8 DESIGN OF SHORT LENGTH CULVERTS UNDER EMBANKMENTS page 21

9 CALCULATING TIME OF CONCENTRATION page 21

10 EXAMPLES OF CALCULATING THE RUNOFF COEFFICIENT C page 22

11 METHOD OF STORMWATER RETICULATION DESIGN page 24

12 GENERAL METHODOLOGY page 24

13 WORKED EXAMPLE page 27

15 CALCULATIONS OF INVERT LEVELS AT MANHOLES page 34

LIST OF APPENDICES page 35

PART R Stormwater Disposal SABS 0400 - 1990 page 36

Revision Date: May 2008 4

Municipal stormwater availability

In many instances erven do not have access or connections to municipal stormwater. In

Therefore, In accordance with the National Building Regulations PART R Stormwater

1.1 Provision of soakpits

In general, where a site was originally developed with attenuation structures/ponds or

Revision Date: May 2008 5

Revision Date: May 2008 6

Soakpits should be constructed 3 4 metres (minimum) from any structure or

1.2 Provision of attenuation and other storage structures

Generally, for multi-unit residential or commercial developments or where coverage

A competent professional civil engineer will be required to produce a Storm-water

The Storm-water Management Plan provided must provide logical and

The pre-development runoff coefficient (C) should ideally be derived on the

The Rainfall statistical data to be used is available on the municipal website:

Revision Date: May 2008 7

Storage or attenuation/infiltration measures must provide for the difference between

The objective is to maximize groundwater infiltration and/or to reduce peak runoff

Revision Date: May 2008 8

1.3 List of minimum information required in a stormwater management plan (for

Street address and description of erf

Revision Date: May 2008 9

2.1 Restriction on Usage

The formula used in this method is

Q = the maximum/peak rate of run-off in cumecs (m3/s)

2.2 Area of catchment (A)

The area of the catchment once defined on a topographic map is determined by

2.3 Time of concentration (Tc)

It is difficult to accurately assess or determine the time of concentration with great

2.4 Run-off coefficient (C)

Revision Date: May 2008 11

Coverall = (Sum of Ci x Ai)/(Sum of Ai) for all of the differing sub-catchments i.

Ci for the catchment i = Csi + Cvi + Cpi

RI Storm year Reduction

Revision Date: May 2008 12