29.1        INTRODUCTION..........................................................................................................29-1

29.2        EROSION AND SCOUR PROTECTION.............................................................................29-1

29.3        ENERGY DISSIPATORS................................................................................................29-2

29.3.1     Riprap Basins for Conduit Outlets.....................................................................29-3

29.3.2     Stilling Basins.................................................................................................29-3

29.3.3     Simple Energy-Dissipating Headwalls...............................................................29-5

29.3.4     Design Criteria and Practices...........................................................................29-5

29.4        DROP STRUCTURES....................................................................................................29-5

29.5        STORMWATER DRAIN OUTFALLS.................................................................................29-10

29.6       TRANSITIONS AND CONSTRICTIONS............................................................................29-13

29.6.1    Transition Analysis.........................................................................................29-13

29.6.2     Constriction Analysis.......................................................................................29-13

29.7        BENDS AND CONFLUENCES.........................................................................................29-15

29.7.1     Bends............................................................................................................29-15

29.7.2     Confluences...................................................................................................29-16

29.8        SIDE-OVERFLOW WEIRS..............................................................................................29-17

29.8.1     Design Considerations....................................................................................29-17

29.8.2     Design Practices.............................................................................................29-17

29.9        FLOW SPLITTER..........................................................................................................29-19

29.9.1     Design Consideration......................................................................................29-20

29.9.2     Design Criteria...............................................................................................29-20

29.10      FLOW SPREADER........................................................................................................29-20

29.10.1   General Design Criteria...................................................................................29-20

29.10.2   Design Criteria for Flow Spreading Options.......................................................29-21

29.11      SIPHONS....................................................................................................................29-24

29.11.1   Single-Barrel Siphons......................................................................................29-25

29.11.2   Multi-Barrel Siphons.......................................................................................29-25

29.11.3   Design Criteria and Practices...........................................................................29-25

Urban Stormwater Management Manual


Hydraulic Structures


Hydraulic structures are used to positively control water flow velocities, directions and depths, the elevation and slope of the stream bed, and the general configuration of a waterway including its stability and maintenance characteristics.

Many of these structures appear as specials and are expensive, which require careful and thorough hydraulic engineering judgement. Proper application of hydraulic structures can reduce initial and future maintenance costs by changing the character of the flow to fit the needs of a particular project, and by reducing the size and cost of related facilities.

The shape, size, and other features of a hydraulic structure can vary widely for different projects, depending upon the functions to be accomplished. Hydraulic design procedures must govern the final design of all structures. This may include model testing when a proposed design requires a configuration that differs significantly from known documented guidelines.


When the flow velocity at a conduit outlet exceeds the maximum permissible velocity for the local soil or channel lining, channel protection is required. This protection usually consists of an erosion resistant reach, such as riprap, between the outlet and the stable downstream channel to provide a stable reach at the outlet in which the exit velocity is reduced to a velocity allowable in the downstream channel. The design of such protection is normally based on a 20 year design runoff event.

If protection is needed at the outlet, a horizontal (zero slope) apron must be provided.

(i) Apron Dimensions

• The length of an apron (La) is determined using the following empirical relationships that were developed for the U.S. Environmental Protection Agency (1976):

4 = ^w + 7Do for TW < Do/2                     (29.1)



La=^mr + 7Do forTW>D0/2                     (29.2)


D0 =              maximum inside culvert diameter (m)

Q =              pipe discharge (cumec)

TW=              tailwater depth (m)

•      Where there is no well defined channel downstream of the apron, the width, W, of the outlet and of the apron (as shown in Figure 29.4) should be as follows:-

W = W0+QALa. forTW> D0/2                       (29.3)


W = 3D0 + 44. for TW < D0/2                       (29.4)

The width of the apron at the culvert outlet should be at least 3 times the culvert width.

•      On the contrary, where there is a well-defined channel downstream of the apron, the bottom width of the apron should be at least equal to the bottom width of the channel and the lining should extend at least one foot above the tailwater elevation and at least two-thirds of the vertical conduit dimension above the invert.

•      The side slopes should be 2:1 or flatter.

•      The bottom grade should be level.

•      There should be an overfall at the end of the apron or culvert.

(ii) Apron Materials

•      The median stone diameter, d50 is determined from the following equation:

O066(«^ 50         TW(D0)                                                   y '

•      Existing scour holes may be used where flat aprons are impractical. Figure 29.5 shows the general design of a scour hole. The stone diameter is determined using the following equatons:

0041^ for 50         TW(D0)                                                   y '

O027(«^ for 50         TW(D0)                        °                         v ;

where Y = depth of scour hole below culvert invert.

•      Other riprap or gabion requirements are as indicated in the previous sections for channel lining.

•      Aprons constructed of man made materials are often a viable alternative. Refer to the above discussion of man-made materials for design consideration.

Urban Stormwater Management Manual


Hydraulic Structures

W = 3Do+0.4La (Tailwater > 0.5Do)

W= 3Do+0.4La (Tailwater < 0.5Do)

Figure 29.1 Configuration of Conduit Outlet Protection (U.S. EPA, 1976)







Figure 29.2 Preformed Scour Hole (ASCE, 1975)


Energy dissipaters are required in the immediate vicinity of hydraulic structures where high impact loads, erosive forces, and severe scour are expected. In other words, they are usually required where the flow regime changes

from supercritical to subcritical, or where the flow is supercritical and the tractive forces or flow velocities are higher than the maximum allowable values. The basic hydraulic parameter that identifies the flow regime, and is used in connection with energy dissipaters in general, and with hydraulic jump dissipaters in particular, is the Froude number (Chapter 12).

The Froude number is a ratio of the flow velocity and wave celerity. In rectangular channels, the equation may be rewritten in the following form:

)2 >




B =    width of channel (m)

Q =    discharge (m3/s)

g =    acceleration due to gravity (9.81 m/s2)

dm =    hydraulic mean depth (m)

Energy dissipation structures act as transitions, which reduce high flow velocities that may exist under a range of flows. Energy dissipaters localise hydraulic jumps and act as stilling basins. The use of energy dissipaters is very common downstream of hydraulic structures where common channel protection cannot be used alone because of potential damage. If riprap or other protection is used


Urban Stormwater Management Manual

Hydraulic Structures

for energy dissipation, it should be confined in a basin and secured in place with grout or mesh.

The primary difference between energy dissipaters in general and drop structures is that the former are used to reduce high velocities at critical locations by hydraulic jumps, while the latter are vertical structures used for controlling velocities in channel reaches by reducing channel slopes. Because of various appurtenances such as sills, baffles, and weirs, and because of variations in the geometry of stilling basins, a wide variety of energy dissipaters may be used.

29.3.1   Riprap Basins for Conduit Outlets

The most commonly used energy dissipaters are riprap basins (Figure 29.3). Their advantages include simplicity, low cost, and wide application.

The riprap placed in the basin must be inspected and repaired, if necessary, after major storms. The median stone diameter can be estimated based on the exit velocity of the pipe or culvert as shown in McLaughlin Water Engineers (1986) and the AASHTO Drainage Handbook (1987). The length of the basin is estimated based on the width or diameter of the conduit. The depth of the basin is based on the median stone diameter.

29.3.2   Stilling Basins

If a hydraulic jump is used for energy dissipation, it should be confined to a heavily-armoured channel reach, the bottom of which is protected by a solid surface such as concrete to resist scouring. Since the cost of concrete structures is relatively high, the length of the hydraulic jump is usually controlled by accessories that not only stabilise the jump action and increase the factor of safety, but also reduce the cost of the structure.

(a) Design Considerations

There are several considerations that should be included in designing hydraulic jumps and stilling basins (Chow, 1959; US DOT, 1983):

1.     Jump Position: There are three positions or alternative patterns that allow a hydraulic jump to form downstream of the transition in the channel. These positions are controlled by tailwater.

2.     Tailwater Conditions: Tailwater fluctuations due to changes in discharge complicate the design procedure. They should be taken into account by classification of tailwater conditions using tailwater and hydraulic jump rating curves.

3.     Jump Types: Various types of hydraulic jumps that may occur are summarised in Figure 29.4. Oscillating jumps in a Froude number range of 2.5 to 4.5 are best avoided unless specially designed wave suppressers are used to reduce wave impact.

The greater the Froude number, the higher is the effect of tailwater on the jump. Therefore, for a Froude number as low as 8, the tailwater depth should be greater than the sequent depth downstream of the jump so that the jump will stay on the apron. When the Froude number is greater than 10, the common stilling basin dissipater may not be as cost-effective as a special bucket type dissipater (see Peterka, 1958).

(b)      Control of Jumps

Jumps can be controlled by several types of appurtenances such as sills, chute blocks and baffle piers. The purpose of a sill located at the end of a stilling basin is to induce jump formation and to control its position under most probable operating conditions. Sharp crested or broad crested weirs can be used to stabilise and control the jump.

Chute blocks are used at the entrance to the stilling basin. Their function is to furrow the incoming jet and lift a portion of it from the floor, producing a shorter length of jump than would occur without them.

Baffle piers are blocks placed in intermediate positions across the basin floor for dissipating energy mostly by direct impact action. They are useful for small structures with low flow velocities. High flow velocities may result in cavitation action on the piers and basin floor downstream.

(c)      Stilling Basin Categories

The following three major categories of basins are used for a range of hydraulic conditions. Design details can be found in the AASHTO Drainage Handbook (1987), Chow (1959), and US DOT (1983).

777e SAF ("St. Anthony Falls" Stilling Basin) ( Chow 1959) : This basin, shown in Figure 29.5, is recommended for use on small structures such as spillways and outlet works where the Froude number varies between 1.7 and 17. The appurtenances used for this dissipater can reduce the length of the basin by approximately 80%. This design has great potential in urban stormwater systems because of its applicability to small structures. Stilling Basin III developed by the US Bureau of Reclamation (UBSR) is similar to the SAF basin, but it has a higher factor of safety.

777e UBSR Stilling Basin II: This basin, shown in Figure 29.6, is recommended for use with jumps with Froude numbers greater than 4.5 at large spillways and channels. This basin may reduce the length of the jump by a third and is used for high-dam and earth-dam spillways. Appurtenances used in this basin include chute blocks at the upstream end of the basin and a dentated still at the downstream end. No baffle piers are used in this basin because of the cavitation potential.

Urban Stormwater Management Manual


Hydraulic Structures

300 mm Free Board Minimum

_ Dissipator Pool _ 10 Ds or 3 W0 Min

Top of Berm Top of Riprap


5hsor W0 Min

Note A

Note B-

3d 50 or 2d Max 1.5 Min

2d so or 2d so or





(a) Centreline Section 6

D ^e^^^Sj

3 Horizontal__

(b) Half-plan




Top of Natural Channel

Thickened or sloping Toe optional Construct if downstream channel degradation is anticipated

-Note B

Symmetrical About Centre Line

Note B-


Sec A-A

-Natural Channel

's ^i^RBo'


Excavate to this Line

2d50 or 1.5dMax

Berm as required to Support Riprap

Sec C-C

Backfill With Riprap

2dso or 1.5d

Sec B-B

Excavate to this line backfill with Riprap

Sec D-D

2d so or 1.5d.v

Berm as required to Support Riprap

Figure 29.3 Typical Riprap Basin: (a) centreline section and (b) half plan: W0 = diameter for pipe culvert, barrel width for box culvert, or span of pipe-arch culvert (US Federal Highway Administration, 1983).

Notes for Figure 29.3:

(1)      If a maximum allowable exit velocity, Ve, from the basin is specified, extend the basin as required to obtain sufficient cross-sectional area at section A-A (i.e. AA.A = Q/Ve) for the specified velocity

(2)      Warp the basin to conform to the natural stream channel. The top of the riprap in the basin floor should be at the same elevation or lower than the natural channel bottom at section A-A


Urban Stormwater Management Manual

Hydraulic Structures

Fj = 1 - 1.7 Undular Jump

F1 = 1.7-2.5 Weak Jump

Oscillating Jet



v • ///v?//........///

F1 = 2.5 - 4.5 Oscillating Jump


Fi > 9.0 Strong Jump

7 S"

F1 = 4.5 - 9.0 Steady Jump

1 5



Undular Weak Oscillating

1/Jump Jump Steady Jump


Strong Jump


Wavy Best Performance Acceptable

-i-i—i ........— Performance

Expensive stilling basin and rough surface conditions

5 I___.......___U___l_l___I___LJ___I___l_l___ I ___LJ___I___I___l_l___I___I___I___LJ___l_l___LJ___ I ___LJ

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Fi = Vi im1

Figure 29.4 Lengths and Types of Hydraulic Jumps in Horizontal Channels (Bradley and Peterka, 1957; Chow, 1959)

• 777e UBSR Stilling Basin IV: This basin, shown in Figure 29.7, is used where jumps are imperfect or where oscillating waves occur with Froude numbers between 2.5 - 4.5. This design reduces excessive waves by eliminating the wave at its source through deflection of directional jets using chute blocks. When a horizontal stilling basin is constructed without appurtenances, the length of the basin is made equal to the length of the jump.

29.3.3   Simple Energy-Dissipating Headwalls

Another simple type of energy dissipater that can be used at culvert outlets is an energy dissipating headwall. Three typical headwalls are shown in Figures 29.8 to 29.10.

29.3.4   Design Criteria and Practices

Most of the design criteria for stilling basin dissipaters are included in the previous paragraphs. Table 29.1 provides a summary of selected parameters, and may be used for

preliminary identification of alternative types of energy dissipaters. Because of the great variety and combination of types of energy dissipaters and appurtenances, the designer should review available references in sufficient detail to arrive at a design that is suitable for specific field conditions.


Vertical drop structures are controlled transitions for energy dissipation in steep channels where riprap or other energy dissipation structures are not as cost effective. Drop structures used for stormwater drainage can be categorised primarily as either open channel transitions (drop spillways) or transitions between open channels and closed conduits (drop shafts).

Drop structures should be constructed of concrete because of the forces involved; however, riprap or gabion stilling basins may be used where physical, economic, and other conditions permit.

Urban Stormwater Management Manual


Hydraulic Structures

(a) Rectangular Stilling Basin Half-plan

Side Wall

Varies Chute Blocks

Floor Blocks



Cut-off Wall

(b) Trapezoidal Stilling Basin Elevation Side Wall

Wing Wall Top Slope

Trapezoidal Stilling Basin Rectangular Stilling Basin Downstream Section

(c) Centreline Section

Figure 29.5 Proportions of the SAF Basin (Chow, 1959).


Urban Stormwater Management Manual

Hydraulic Structures

Dentated Sill

Slope = 2:1

Figure 29.6 Proportions of the USBR Basin II (Chow, 1959)

Fractional Space

w = Maximum Tooth Width D! Space = 2.5w

Top Surface on 5-deg slope 2D!

Sill Optional

Figure 29.7 Proportions of the USBR Basin IV (Chow, 1959)


300 mm'300 mm'300 mm R8 Re-bar

300 mm Square Solid Concrete Block Formed -

and Poured in Place, Reinforced by

R8 Re-bars Dispersed Every 100 mm Throughout

300 mtfpOO mmpOO mm|

100 mm

(a) Plan

(b) Side Elevation

Figure 29.8 Standard Energy Dissipating Headwall Type I (Chow, 1959)

Urban Stormwater Management Manual


Hydraulic Structures


Height of Energy Dissipators should be 0.5


Figure 29.9 Standard Energy Dissipating Headwall, Type II (ASCE, 1992)



R8 at 300 mm Centres

(a) Front Elevation

R8 200 mm Centres

Precast Concrete

25 mm Minimum

450 mm

(b) Side Elevation

Figure 29.10 Standard Energy Dissipating Headwall, Type III (ASCE, 1992)


Urban Stormwater Management Manual

Hydraulic Structures

Table 29.1 Dissipater Criteria (U.S. Department of Transportation, 1983)

Allowable Debris

Dissipater Type

Froude Number F

Tailwater TW

Special Considerations

Silt and Sand



Free Hydraulic Jump

> 1





CSU Rigid Boundary

< 3





Tumbling Flow

> 1





4% < S0< 25%

Increased Resistance






Check Outlet Control HW


4 to 14






4.5 to 17






2.5 to 4.5






1.7 to 17





Contra Cost







1.8 to 3











Q<11 m3/s, V< 15 m/s

Forest Service






y< 900 mm

Drop Structure

< 1





Drop < 5 m







Corps Stilling Well












Note: N = none, L = low, M = moderate, H = heavy

Drop structures in open channels change the channel slope from steep to mild by combining a series of gentle slopes and vertical drops. Flow velocities are reduced to non-erosive velocities, while the kinetic energy or flow velocity gained by the water as it drops over the crest of each spillway is dissipated by an apron or stilling basin.

Open channel drop structures generally require aerated nappes and subcritical flow conditions at both the upstream and downstream section of the drop. The stilling basin can vary from a simple concrete apron to baffle blocks or sills as described previously.

Figure 29.11 shows the flow geometry and important variables at a vertical (straight) drop structure. The flow geometry at such drops can be described by the drop number, DNl which is defined (Chow 1959) as:

DN = -^j                                                     (29.9)


q = discharge per unit width of crest overfall (m3/s/m) g = acceleration due to gravity (9.81 m/s2) h = height of drop (m)

The drop functions are:

^ = 4.30 V-27 h


^- = 1.00 DN022 h


^ = 0.54Z?/-425


^ = 1.66Z?/-425



Ld = drop length (m)

yp = pool depth under the

nappe (m)

Urban Stormwater Management Manual


Hydraulic Structures

/! = the depth of the toe of nappe (m) y2 = tailwater depth sequent to /i (m)

For a given drop height, h, and discharge, q, the drop length, Ldl and the sequent depth, y2, can be estimated by Equations 29.10(a) and 29.10(d), respectively. The length of the jump can be estimated by techniques discussed in Section 29.3. If the tailwater is less than y2, the hydraulic jump will recede downstream. Conversely, if the tailwater is greater than y2, the jump will be submerged. If the tailwater is equal to y2, no supercritical flow exists on the apron and the distance Ld\s minimum.

When the tailwater depth is less than y2l it is necessary (according to the US Department of Transportation 1983) to provide either (1) an apron at the bed level and a sill or baffles, or (2) an apron below the downstream bed level and an end sill.

The choice of a design type and dimensions depends on the unit discharge, q, drop height, h, and tailwater depth, TW. The design should take into consideration the geometry of the undisturbed flow. If the spillway (overflow crest) length is less than the width of the approach channel, the approach channel must be designed properly to reduce the effect of the end contractions to avoid scour.

The two most common vertical open channel drops are the straight drop structure and the box inlet drop structure.

(a) Straight Drop Structure

Figure 29.12 shows the layout of a typical straight drop structure and hydraulic design criteria developed by US Soil Conservation Service. McLaughlin Water Engineers (1983) provides specific criteria and reviews design considerations related to the hydraulic, geotechnical, and structural design of drop structures.

(b) Box Inlet Drop Structure

The box inlet drop structure is a rectangular box open at the top and downstream end as shown in Figure 29.13. Water is directed to the crest of the box inlet by earth dikes and a headwall. Flow enters over the upstream end and two sides. The long crest of the box inlet permits large flows to pass at relatively low heads. The width of the structure should not be greater than the downstream channel. Box inlet drop structures are applicable to drops from 0.6 to 3.6 m.

Design data and criteria for these structures, based on US Soil Conservation Services and St. Anthony Falls Hydraulic Laboratory, are available (US Department of Transportation 1983; Blaisdell and Donnely 1956). The parameters to consider for the hydraulic design of the drop are:

•      section (length) of the crease of the box inlet

•      opening of the headwalls

•      discharge, discharge coefficients, and flow regime changes

•      box inlet length and depth

•      minimum length and width of stilling basin


All stormwater drains of a locality have an outlet where flow from the local drainage system is discharged. The discharge point, or outfall, can be either a natural river or stream, or an existing or proposed stormwater drain or channel. The procedure for calculating the hydraulic grade line through a storm drainage system begins at the outfall. Therefore, consideration of the outfall conditions is an important part of storm drain design.


Figure 29.11         Flow Geometry of a Straight Drop Spillway (Chow, 1959)


Urban Stormwater Management Manual

Hydraulic Structures

(1200 mm Minimum) T

150 mm Fillets

(a) Section on Centreline

F + S

C (1200 mm Minimum)

150 mm

(b) Downstream Elevation

Level Area

(c) Plan

E = Minimum length of headwall extension = [3h+0.61] or [1.5F] whichever is greater

J = Height of wing wall and sidewall at junction = [2h] or |"p+[-|+s_/,LB+ 0.13A"| or [t+1] whichever is greater                                                    L            v j 'J

LB= Length of basin = |> (2.28 £- + 0. 52^ ]

M = [2 (F+ 1.3 h-J)] K = [(LB+0.13)-M]

(d) Criteria

Figure 29.12 Typical Drop Spillway and Some Hydraulic Design Criteria (US Soil Conservation Service, 1954).

Urban Stormwater Management Manual


Hydraulic Structures

Toe of Dike


60° > 6 > 45°

(a) Plan


(b) Section on Centreline

Figure 29.13 Box Inlet Drop Structure (US Dept. of Transportation, 1983)

Several aspects of outfall design must be given serious consideration. These include the flowline or invert (inside bottom) elevation of the proposed stormwater drain outlet, tailwater elevations, the need for energy dissipation, and the orientation of the outlet structure.

The flowline or invert elevation of the proposed outlet should be equal to or higher than the flowline of the outfall. If this is not the case, there may be a need to pump or otherwise lift the water to the elevation of the outfall (see Chapter 46).

The tailwater depth or elevation in the storm drain outfall must be considered carefully. Evaluation of the hydraulic grade line for a storm drainage system begins at the system outfall with the tailwater elevation. For most design applications, the tailwater will either be above the crown of the outlet or can be considered to be between the crown and critical depth of the outlet. The tailwater may also occur between the critical depth and the invert of the outlet; however, the starting point for the hydraulic grade line determination should be either the design tailwater elevation, or {dc + D)/2, whichever is higher.

An exception to the above rule would be for a very large outfall with low tailwater where a water surface profile calculation would be required to determine the location where the water surface will intersect the top of the barrel and full flow calculations can begin. In this case, the downstream water surface elevation would be based on critical depth or the design tailwater elevation, whichever was higher.

If the outfall channel is a river or stream, it may be necessary to consider the joint or coincidental probability of two hydrologic events occurring at the same time to adequately determine the elevation of the tailwater in the receiving stream (see chapter 46.7.2 for details).

Energy dissipation may be required to protect the storm drain outlet. This is to prevent erosion of the outfall bed and banks. Riprap aprons or energy dissipaters should be provided if high velocities are expected.

There may be instances in which an excessive tailwater causes flow to back up the storm drainage system and out of inlets and manholes, creating unexpected and perhaps hazardous flooding conditions. The potential for this to occur should be considered. Flap gates placed at the


Urban Stormwater Management Manual

Hydraulic Structures

outlet can sometimes alleviate this condition (see chapter 46); otherwise, it may be necessary to isolate the storm drain from the outfall by use of a pump station (see Chapter 46).

The orientation of the outfall is another important design consideration. Where practical, the outlet of the storm drain should be positioned in the outfall channel so that it is pointed in a downstream direction. This will reduce turbulence and the potential for excessive erosion. If the outfall structure cannot be oriented in a downstream direction, the potential for outlet scour must be considered. For example, where a storm drain outfall discharges perpendicular to the direction of flow of the receiving channel, care must be taken to avoid erosion on the opposite channel bank. If erosion potential exists, a channel bank lining of riprap or other suitable material should be installed on the bank. Alternatively, an energy dissipater structure could be used at the storm drain outlet.


Channel transitions (Figure 29.17) are typically used to alter the cross-sectional geometry, to allow the waterway to fit within a more confined right-of-way, or to purposely accelerate the flow to be carried by a specialised high velocity conveyance. Constrictions can appreciably restrict and reduce the conveyance in a manner which is either detrimental or beneficial. For example, a bridge, box culvert, or other constriction may increase upstream flooding by encroaching too far into the floodplain conveyance, whereas in another situation, a hydraulic control structure can be employed to purposely induce an upstream spill into a storage facility.

The purpose of this section is to briefly outline typical design procedures for transition and constriction structures that may be required for engineered waterways.

29.6.1 Transition Analysis

(a) Subcritical Transitions

Transitions for subcritical flow frequently involve localised or bank lining configurations which allow change in the cross section and produce a water surface profile based on gradually varied flow. The energy lost through a transition is a function of the friction, eddy currents, and turbulence. The intent is often to minimise friction losses and/or The intent is often to minimise friction lossess and/or erosion tendencies. Examples include transitions between trapezoidal and rectangular sections, modest transitions at bridges where little change takes place in cross section, or slight encroachments into a channel to allow for utilities.

Standard water surface profile analysis is applied, with the addition of an energy loss at the transition. The loss is

expressed as a function of the change in velocity head occurring across the contraction or expansion transition (from upstream to downstream locations). Figure 29.14 illustrates some of these transitions. Loss coefficients shown in Table 29.7 are applied to the difference in the velocity head, as shown is Equation 29.11.

Table 29.2 Subcritical Transition Energy Loss Coefficients



Less than 100 mm between

centreline and tangent lines



Less than 12.5° between

0 to 0.10

0 to 0.10

centreline and tangent lines

Warped Type



Cylindrical Quadrant Type



Modest Transitions



Straight Line Type



Square Ended Type



Analysis of transitions requires careful water surface profile analysis including verification of effective channel hydraulic controls. It is not uncommon to have a transition which is first thought to be performing in a subcritical mode, but subsequently found to produce a supercritical profile with a hydraulic jump.

Energy Loss (m) = Coefficient \hvl - hv2) (29.11) where,

\ vl       vl! y2g lg

Vx = flow velocity upstream of transition V2 = flow velocity downstream of transition

(b) Supercritical Transition Analysis

Supercritical transitions are beyond the scope of this manual and require special analysis when used. The configuration of a supercritical transition is entirely different from subcritical transitions. Improperly designed and configured supercritical transitions can produce shock wave patterns which result in channel overtopping and other hydraulic and structural problems.

29.6.2 Constriction Analysis

(a) Constriction with Upstream Subcritical Flow

There are a variety of structures that are constrictions. They can include bridges, culverts, drop structures, and flow measurement devices. Constrictions of various types are used intentionally to control bed stability and upstream

Urban Stormwater Management Manual


Hydraulic Structures

water surface profiles. For example, a constriction may be used to cause water to back up or overflow into a detention basin.

The hydraulic distinction of constrictions is that they can cause rapidly varied flow. The upstream transition loss coefficients in Table 27.9 apply, but other factors come into play. Significant eddies can form upstream and downstream of the constriction depending upon the geometry. Flow separation will start at the upstream edge of the constriction, then the flow contracts to be narrower than the opening width. Typically, the width of contraction is 10% of the depth at the constriction for each side boundary. For example, at a typical drop with an abrupt crest contraction and assuming critical depth of 1.0 m, the constriction on each side would be 100 mm, or 200 mm total contraction from the opening width. Based on this contracted width and an assumption of critical conditions

at that location, the upstream water surface profile may be computed.

In certain cases, the flow regime will remain subcritical through the constriction. Chow, 1959, presents guidelines developed by the USGS for constrictions where the Froude number in the contracted section does not exceed 0.8. These cases are generally mild constrictions.

A phenomenon of abrupt contractions (and abrupt expansions) is that the velocities can be much higher in the centre and change significantly across the constriction throat section. This results in a large energy coefficient and a further drop in water surface over what is first anticipated. This condition can produce strong eddy currents with a high erosion potential. A constriction in an open channel needs to be carefully evaluated for velocity, scour, water surface, and related problems.

Cylindrical Quardrant

Straight Line





Figure 29.14 Transition Types (UDFCD, 1969)


Urban Stormwater Management Manual

Hydraulic Structures

Constrictions used for flow depth control or flow measurement devices require a high degree of accuracy. The design information available that can be used for ensuring a high degree of accuracy is limited. It is advisable to use models tested or proven prototype layouts. As a secondary option, adjustable edge plates or other components can be provided to allow later changes at minimal cost if the constructed facilities should need refinement.

(b) Constriction with Upstream Supercritical Flow

upstream and in accelerated velocity zones, with high possibility of erosion on the outside of the bend and other locations. Significant eddy currents, scour, sedimentation, and loss of effective conveyance can occur on the inside of the bend.

Concrete lined channels can be significantly affected by superelevation of the water surface. The designer should always add superelevation to the design freeboard of the channel. The equation for the amount of superelevation of the water surface, Ay, that takes place is given as:

This situation is highly complex and beyond the scope of this manual. Possible shock waves or choked flow causing high upstream backwater or a hydraulic jump are major concerns. The situation is to be avoided in urban drainage because of inherent instabilities.


General considerations for lined channels and conduits are discussed in Chapters 26 and 25. Additional emphasis is added herein for certain situations. Channels and conduits that produce supercritical flow may require special structural or design considerations. This discussion is limited since these types of structures are associated with hydraulic performance that generally exceeds the recommended criteria for most engineered waterways. Extensive study, specialised modelling, and/or analysis may be required for these situations.

Channel confluences are commonly encountered in design. Flow rates can vary disproportionately with time so that high flows from upstream channel can discharge into downstream channel when it is at high or low level. Depending on the geometry of the confluence, either condition can have important consequences, such as supercritical flow and hydraulic jump conditions, and result in the need for structures.

The main emphasis in this section is on subcritical flow conditions. Since supercritical conditions can occur in various situations, some conditions are generally reviewed; however, supercritical flow analysis is not described in detail.

29.7.1 Bends

(a) Subcritical Bends

Subcritical bends are required to have certain minimum curvatures described in Chapters 25 and 26. It is important that the designer recognise the consequences of approaching and exceeding these criteria. Chow, 1959, Rouse, 1949, and others illustrate flow patterns, superelevation, and backwater or flow resistance characteristics. Superelevation refers to a rise in the water surface on the outer side of the bend. Effectively, the bend can behave like a contraction, causing backwater

Ay = C









coefficient, generally 0.5 for subcritical flow mean channel velocity (m/s) width of water surface in channel (m) acceleration of gravity (9.81 m/s2) channel centreline radius (m)

(b) Supercritical Bends

As with supercritical transitions, the hydraulics for supercritical bends are completely different from the hydraulics for subcritical bends. Supercritical channels are generally not desirable in urban drainage. However, special situations may occur where supercritical flows enter a curved channel, for example:

•      at confluences where one channel is largely empty, and the entering flow expands and becomes supercritical

•      at a sharp bend in a conduit whose slope inherently leads to supercritical conditions

•      at a channel drop that unavoidably ends up on a curve

The key phenomenon to be aware of is shock waves, of which there are two types, positive and negative. On the outside of an angular bend, a positive shock wave will occur which results in a rise in the water surface. The wave is stationary and crosses to the inside of the channel, and then can continue to reflect back and forth. Where the flow passes the inside angular bend, a separation will occur and a negative shock wave or drop in the water surface will occur. This stationary negative shock wave will cross to the outside of the channel. Both shock waves will continue to reflect off the walls, resulting in a very disturbed flow pattern.

A basic control technique is to set up bend geometry to cause the positive shock wave to intersect the point where the negative wave is propagated. A bend usually requires two deflections on the outside and one bend on the inside. A beneficial aspect of the shock wave is that it turns the

Urban Stormwater Management Manual


Hydraulic Structures

flow in a predictable pattern, and thus the channel walls have no more force imposed on them other than that caused by the increased (or decreased) depths. This technique is described by Ippen (Rouse, 1949 and Corps of Engineers, 1970) and is also reported by Chow, 1959.

Other control techniques include very gradual bends, super elevated floors, and controls sills, but these methods are generally less efficient. There is limited data on channels with sloping side banks, but it is clear there is a great tendency for shock waves to propagate up side slopes and divert flow out of the channel.

Chow, 1959 shows several good photographs of these problems. SCS, 1976 presents a documental report of a curved spillway on a modest flood control storage. During an overflow event, a shock wave pattern was produced that resulted in no flow on one side of the spillway and great depths on the opposite.

A special problem with long conduits used for flood control, particularly large box culverts, is that they will have an inherent tendency towards supercritical flow conditions at less than full capacity. When the flow encounters bends, shock waves can occur which hit the ceiling and can lead to either pressurised conditions and/or unstable conditions where the flow fluctuates between supercritical and pressure conditions, often exacerbated by surging, air flow, and other problems. Very gradual bends and air vent provisions are desirable and the designer should use caution anytime should supercritical flow encounter a bend.

29.7.2 Confluences

One of the most difficult problems to deal with is confluences where the difference in flow characteristics may be great. When entering the combined channel, the flow can diverge and drop in level if the flow capacity is suddenly increased. This can result in high velocity or unstable supercritical flow conditions with a high erosion potential. When significant sediment flows exist, aggradation can occur at the confluence, resulting in the loss of capacity in one or both upstream channels.

(a) Subcritical Flow Confluence Design

The design of channel junctions is complicated by many variables such as the angle of intersection, shape and width of the channels, flow rates, and type of flow. The design of large complex junctions should be verified by model tests.

Figure 29.15 illustrates two types of junctions. The following assumptions are made for combining subcritical flows:

1. The side channel cross-section is the same shape as the main channel cross-section


2.     The bottom slopes are equal for the main channel and side channel

3.     Flows are parallel to the channel walls immediately above and below the junction

4.     The depths are equal immediately above the junction in both the side and main channel

5.     The velocity is uniform over the cross-sections immediately above and below the junction

Assumption No. 3 implies that hydrostatic pressure distributions can be assumed, and assumption No. 5 suggests that the momentum correction factors be equal to each other at the reference sections.

The equation governing flow conditions for a vertical walled channel, with the main channel width being constant, is (Figure 29.15(a)):

-9L+b y3=-9L+cos 0-9L+hzL (29.i3a)

gA3               gA1           g A2 2

Or for a vertical walled channel; main channel width varies (Figure 29.15(b)):

-OL + by2 -QL +cos0 J?L + ?3lL           (29.13b)

gA3              gA1           g A2 2

Or for a trapezoidal channel; main channel width constant (Figure 29.15(a)):

gA3 [2 3 y3 gA,           gA2 [2 3 J'1


Or for a trapezoidal channel; main channel width varies (Figure 29.15(b)):

gA3 [2 3 y3 gA,           g A2 { 2 3 J'1

(29.13d) where,

b = bottom width of the trapezoidal cross-section Z = side slope, Z(H):1(V)

Momentum computations for a confluence involve a trial and error process. Starting with a known depth above or below the confluence, one iterates with an assumed depth on the unknown side of the confluence until the momentum has been balanced upstream to downstream

(b) Supercritical Flow in Confluences

In contrast with subcritical flows at junctions, supercritical flows with changes in boundary alignments are generally complicated by standing waves (Ippen, 1951, Rouse,

Urban Stormwater Management Manual

Hydraulic Structures

1949). In subcritical flow, backwater effects are propagated upstream thereby tending to equalise the flow depths in the main and side channels. However, backwater cannot be propagated upstream in supercritical flow and flow depths in the main and side channels cannot generally be expected to be equal. Junctions for rapid flows and very small junction angles are designed assuming equal water surface elevations in the side and main channels.

Standing waves (Ippen, 1951) in supercritical flow at open channel junctions complicate flow conditions. These waves may necessitate increased wall heights in the vicinity of the junction. The studies by Bowers, 1950 indicate that a hydraulic jump may form in one or both of the inlet channels, depending on the flow conditions. Behike, 1966 completed a model study and presented design guidelines to address these problems.

Wave conditions that may be produced by rapid flow in and downstream of a typical junction are shown in Figure 29.19. One area of maximum wave height can occur on the side channel wall opposite the junction point and another on the main channel right wall downstream from the junction. Behike, 1966 has conducted a series of laboratory tests which indicate that wave pileup against the channel walls can be up to seven times the initial depth for a flow Froude number of four. The design of walls to contain these wave heights over long channel distances is usually not economical. The practical remedy is to reduce or minimise standing waves.

Peak flows from the side channel may not occur simultaneously with peak flows in the main channel. Laboratory tests by Behike, 1966 indicate that the occurrence of design flow in one of the channels with zero flow in the other can result in a very high wave pileup on the junction walls.

Supercritical flow may unavoidably occur in certain confluences. The designer should try to correct the geometry and channel sections to avoid the situation. If the condition remains, a more detailed hydraulic study or model study must be initiated to address the problem.


Side-overflow weirs facilitate overflow and diversion of stormwater by directing the discharge away from the original channel. Such structures are commonly used to direct channel discharges above predetermined levels into off-line stormwater detention facilities. Flow diversions occur only during storms (Figure 29.17).

29.8.1 Design Considerations

The design of side-overflow weirs is based on empirical equations which quantify the relationship between the

discharge over the weir and geometric parameters at the weir, including the length of the weir and head (Hager, 1987). Figure 29.18 (Metcalf and Eddy, 1972) shows three head or water surface profile conditions that can prevail at a side-overflow weir:

(a)   Condition 1: The channel bed slopes steeply, producing supercritical flow. Under this condition, the weir has no effect upstream and along the weir there is a gradual reduction in depth. Downstream of the weir, the flow depth in the original channel increases, tending asymptotically to the normal depth corresponding to the remaining discharge.

(b)   Condition 2: The channel bed slopes mildly. Under this condition, subcritical flow prevails and the weir impact is noticed upstream of the weir only. The water surface profile downstream of the weir corresponds to the normal depth of the remaining discharge. Along the weir there is a gradual increase in depth and, upstream of the weir the flow depth tends asymptotically to the normal depth for the initial discharge.

(c)   Condition 3: The channel bed slopes mildly, but the weir crest is below the critical depth corresponding to the initial flow, and the flow at the weir is supercritical. Recent studies (Frazer 1957) indicate that conditions 1 and 3 may result in the development of a hydraulic jump at the weir.

The most common condition that a designer will encounter is Condition 3, where the weir elevation is below the critical depth. When only a relatively small amount of the flow is diverted, a rising water surface profile occurs. According to Metcalf and Eddy Inc. (1972), the falling profile results if the ratio of the height of the weir, c, to the channel specific energy, Ew, referenced to the top of the weir, is less than 0.6.

29.8.2 Design Practices

(a) Falling Water Surface

The equations and procedures for computing weir length for the falling water surface profile were developed by Ackers (Chow 1959). These equations combine Bernoulli's theorem with a weir discharge formula. Metcalf and Eddy Inc. (1972) suggests using:

L = 2.035 5.28-2.63—                             (29.14)


L =    length of weir (m)

B =    channel width (m)

c =    height of weir (m)

Ew =    channel specific energy (m)


Urban Stormwater Management Manual


Hydraulic Structures

(a) Plan - Constant Width


(b) Plan - Unequal Width Figure 29.15 Channel Junction Definition Sketches

Area of Maximum Wave Height

Area of Maximum

Wave Height When B2>9°


0 Should Not be Greater than 12°

Figure 29.16 Open Channel Confluence, Standing Waves - Supercritical Flow


Urban Stormwater Management Manual

Hydraulic Structures





a    =    velocity coefficient

V    =    normal velocity in the approach channel (m/s)

g    =    acceleration due to gravity(9.81 m/s2)

a'   =    pressure-head correction

yn = normal depth of flow in approach channel (m) c = height of the weir above the channel bottom (m)

Values for a and a' of 1.2 and 1.0 respectively can be

used in the approach channel, while at the lower end of the weir values of 1.4 and 0.95 can be used for a and a' respectively.

(b) Rising Water Surface

The analysis for estimating the weir length for the rising water surface profile is based on the theoretical equations developed by DeMarchi (Collinge 1957):







length of weir (m) channel width (m) constant (0.35 for a free nappe)

varied flow function (Figure 3, Collinge 1957)

depth in channel (m) : specific energy (m)

Equation 29.16 is recommended for use only in the case of a rising water surface profile. Metcalf and Eddy Inc. (1972) indicates that this equation works best when the Froude number is between + 0.3 - 0.92.


A flow splitter is a special structure designed to divide a single flow and divert the parts into two or more downstream channels. A flow splitter can serve three functions.


Reduction in water surface elevation - By dividing the flow from a large pipe into multiple conduits, the height of flow (measured from the flow line to the water surface (or for pipes flowing full, the inside diameter) can be reduced. This may be necessary to route flows under immovable obstructions.

Energy Grade Line



Figure 29.17 Typical Cross-sections at a Side-Overflow Weir (Metcalf & Eddy, 1972)

Urban Stormwater Management Manual


Hydraulic Structures

(a) Condition 1

(b) Condition 2

C > dcrit and L -             c~"'

(c) Condition 3

Figure 29.18 Possible Types of Water Surface Profiles at a Side-overflow Weir (Metcalf & Eddy, 1972).

(b)    Dividing flows wherever necessary - Examples of this include division of existing large special-design conduits, such as arches or horseshoes, into less expensive multiple-pipe continuations and division of flow between low and high-flow conduits at the intake of an inverted siphon.

(c)    To restrict flows to water quality treatment facilities and bypass the remaining higher flows around the facilities (off-line). This can be accomplished by splitting flows in excess of the water quality design flow upstream of the facility and diverting higher flows to a bypass pipe or channel. The bypass typically enters a detention pond or the downstream receiving drainage system. A crucial factor in designing flow splitters is to ensure that low flows are delivered to the treatment facility up to the water quality design flow rate. Above this rate, additional flows are diverted to the bypass system with minimal increase in head at the flow splitting structure to avoid surcharging the water quality facility under high flow conditions.

Figure 29.19 shows a typical flow splitter made of manholes with concrete baffles. Figure 29.20 shows a typical diversion/isolation structure.

29.9.1   Design Consideration

Two major considerations exist for the design of flow-splitting devices:

(a)   Head Loss - Hydraulic disturbances at the point of flow division result in unavoidable head losses. These losses, however, may be reduced by the inclusion of proper flow deflectors in the design of the structure. Deflectors minimise flow separation by providing a gradual transition for the flow, rather than by forcing abrupt changes inflow direction.

(b)   Debris - In all transitions from large to smaller pipes, debris accumulation is a potential problem. Tree limbs and other debris that flow freely in the larger pipe may not fit in the smaller pipe(s) and may restrict flow. In addition, flow splitters cause major flow disturbances resulting in the regions of decreased velocity. This reduction causes material suspended in the stormwater flow to settle in the splitter box. Although the deflector design should minimise velocity reduction as much as possible, total elimination of the problem is unlikely. Therefore, positive maintenance access must be provided. Because flow splitting devices are maintenance-intensive, their use should be judiciously controlled by the engineer.

29.9.2   Design Criteria

•      The flow splitter shall be designed to cater for 6 month ARI storm.

•      The top of the weir shall be located at the water surface for the 3-month ARI water quality design storm.

•      The maximum head over the weir shall be minimised for flow in excess of the water quality design flows.


Flow spreaders are used to uniformly spread flows across the inflow portion of water quality facility (e.g. sand filter, biofiltration swale, or filter strip). Options A through C (see Section 29.10.2) can be used for spreading flows that have already concentrated. Option D is only for flows that are already unconcentrated and enter a filter strip or biofiltration swale.

29.10.1 General Design Criteria

•      Where flow enters the spreader through a pipe, it is recommended that the pipe be submerged to practically dissipate energy.

•      Rock protection is required at outfalls.


Urban Stormwater Management Manual

Hydraulic Structures

29.10.2 Design Criteria for Flow Spreading Options

The following presents the design criteria for each of the following spreading options:

•      Anchored plate (Option A)

•      Concrete sump box (Option B)

•      Flat-topped notched curb spreader (Option C)

•      Through-curb ports (Option D)

(a) Option A - Anchored Plate

Figure 29.21 shows the details of the spreader.

The spreader shall be preceded by a sump having a minimum depth of 200 mm and minimum width of 600 mm. The sump area shall be lined with steps to reduce erosion and to provide energy dissipation.

The top of the flow spreader plate shall be level, projecting a minimum of 50 mm above the final grade of the invert of the water quality facility.

The plate shall extend horizontally beyond the bottom width of the facility to prevent water from eroding the side slope.

The plate shall be securely fixed in place.

The level spreader plate may be either wood, metal, fibreglass reinforced plastic, or other durable material.


Type 2 M.H No Base Channel Required



To Bypass Conveyance System or Detention Pond

Reinforced Baffle Wall Grouted to M.H Structure (both ends)

To Water Quality Facility

Round Solid Lid


Water Quality Design Water Surface Elevation

Hands Holds (typ.)



1220 mm or Provide Separate Access to Either Side of Baffle Wall

100 mm Thick Reinforced Concrete Baffle Wall or Other Suitable Material

Bypass Pipe

To Water Quality Facility

Sump (Optional)


The water quality discharge Pipe may require an orifice plate be installed on the Outlet to control the height of the design water surface (weir height). The design water surface should be set to provide a minimum headwater/diameter ratio of 2.0 on the outlet Pipe.

Figure 29.19 Typical Flow Splitter Device

Urban Stormwater Management Manual


Hydraulic Structures



Diversion Weir

Open Channel







Diversion Pipe Diameter, D

► A

Diversion Weir

Open Channel

to Filtration Basin


Figure 29.20 Typical Isolation/Diversion Structure

(b)      Option B - Concrete Sump Box

This alternative uses a rectangular concrete sump (see Figure 29.22 for details)

•      The wall of the downstream side of the concrete sump shall extend a minimum 50 mm above the invert of the treatment bed.

•      The downstream wall of the box shall have "returns" at both ends. Side walls and returns shall be slightly higher than the weir so that erosion of the side slope is minimised.

(c)      Option C - Flat- Topped Notched Curb Spreader

Flat-topped notched curb spreader is shown in Figure 29.23. The spreader sections are made of extruded concrete laid side by side and level. Typically four "teeth"

per 1.25 m section provide good spacing. The space between adjacent "teeth" forms a v-notch.

(d) Option D - Through Curb

Details of the spreader are shown in Figure 29.24. Unconcentrated flows from paved areas entering filter strips or continuous flow biofiltration swales can use curb ports to allow flows to enter the strip or swale. Curb ports use prefabricated openings that allow concrete curbing to be poured or extruded continuously while still providing an opening through the curb to admit water to the water quality facility.

Openings in the curbing shall be at regular intervals but at least every 3.6 (minimum). The width of each curb port opening shall be 275 mm minimum. Approximately 15 percent or more of the curb section length shall be in open ports, and no port should discharge more than about 10 percent of flow.


Urban Stormwater Management Manual

Hydraulic Structures


Example of anchored plate used with a sand filter* (May also be used with other water quality facilities)

V-Notched or

Level Plate Spreader

Extended into slope to protect from the 100 year flow or the highest flow entering water quality facility

Anchor posts spaced 1.8 m O.C or at each end if width 1.8 m

Edge of Sand Pond Side Slope

*Sand filter may use other spreading options


Rock Rip Rap Pipe

200 mm

Sand Layer Gravel Layer

Alternative Design

Example of Catch Basin recommended for higher flow situations (Generally 140 L/S or greater for 100 year storm)

Existing Grade

Level Spreader Plate bolted to anchor Post

600 mm Embedded into exisitng ground

Figure 29.21 Level Spreader (Option A)

Urban Stormwater Management Manual


Hydraulic Structures


Example of a Concrete Sump Flow Spreader

used with a biofiltration swale

(May be used with other W.Q. Facilities)

Concrete Sump

Outfall Rip Rap Pad

Side with Returns. (See Note)


Inlet Pipe

Wing Wall Outline

50 mm Clearance


Concrete Sump

(100 mm wall thickness)


Extend sides into slope. Height of side wall and returns must be sufficient to handle the 100 year flow or the highest flow entering the Facility

Figure 29.22 Level Spreader (Option B)


Any conduit that drops under an obstruction such as railroad tracks, depressed roadways or utilities, and regains elevation at the downstream side of the obstruction is referred to as an inverted siphon.

Because of the inverted bottom, the siphon stands full storm water even when there is no flow. Some drainage districts discourage the use of siphons on the basis that the siphon requires more frequent maintenance including removal of debris that may clog the conduit. Nevertheless, siphons have certain advantages in particular settings, usually in urban areas where other solutions such as flow


Urban Stormwater Management Manual

Hydraulic Structures

re-routing may result in disruptions to traffic and higher costs.

Siphons are normally single-or multi-barrel and consist of an entrance section, drop, depressed reach, rise, and outlet structure. Siphons require hydraulic head to operate properly and the adequacy of available head should be assessed early in the design process. Siphons can be simpler or sophisticated, and the related design effort can be nominal or complex. The following examples apply to large, sophisticated siphons with multiple appurtenances -some or which may not always be necessary.

29.11.1     Single-Barrel Siphons

Single-barrel siphons can be used for conveying stormwater flows where there are periods of no flow during which maintenance can be provided.

Even though some agencies limit the slope of the rising leg of the siphon to 15%, steeper slopes and even vertical drops and risers are acceptable, if maintenance chambers with debris collection sumps at the bottom are provided at the drop and riser of the siphon, as shown in Figure 29.25.

Sloping legs of siphons (Figure 29.26) have been designed without maintenance chambers; however, the chambers provide maintenance flexibility with direct access to service the siphon. The steeper the legs of the siphon, the more difficult it is to clean the siphon from shallow manholes located near the ground surface, and deep maintenance chambers reaching to barrel inverts may be required.

Where a vertical drop and riser are provided, they should serve as maintenance chambers and include access down to the barrels and sumps. Sumps located at the bottom of the maintenance chambers trap the debris that accumulates in the siphon.

29.11.2     Multi-Barrel Siphons

In channels or sewers that convey a continuous flow, where one barrel does not have sufficient capacity and the flow has to be divided, or where redundancy is required by local agencies, the multi-barrel siphon is applicable. Plan and profiles of such siphons are shown in Figure 29.26.

Where redundancy is required for maintenance purposes, one additional equal capacity barrel is sufficient. To fulfil its functions, the multi-barrel siphon requires equipment and structure , including gates that close the barrel to be maintained while the other barrel is open.

Special structure may also include a flow distribution chamber and a flow adapter chamber. These chambers are used to contract and expand the flows. The distribution chamber serves to direct the flow from one sewer to the two barrels of the siphon alternatively used, while the flow adapter chamber serves to direct the flow from the two barrels of the siphon to one conduit.

29.11.3 Design Criteria and Practices

One of the critical criteria for the design of siphons is the maintenance of self-cleansing velocities under widely varying flow conditions (ASCE 1969). Siphons used for conveying storm water are usually designed for a velocity of 0.9 m/s for a 5-year return interval design flow. Siphons with water containing abrasive suspended materials should be designed for a flow velocity less than 3 m/s.

The head losses through each of the siphon components must be estimated for the purpose of plotting the hydraulic grade line. Upstream surcharging should be avoided, and therefore one of the main design objectives should be to minimise the head losses through the siphon. The friction losses can be estimated by using the combined Darcy-Weisbach - Manning equation is useful in the following form (in metric units):

. 19.5 n2LV2                                                 ,-,01-n

hr= r<l32g                                                               (29-1?)


hf = lead loss (m)

n = Manning's frictions factor

L = length of conduit (m)

r = hydraulic radius (m)

V = velocity (m/s)

g = acceleration of gravity (m/s2)

Minor losses (such as at bend, contraction and expansion, and entrance and exit losses) can be estimated as discussed in Chapter 25. It should be noted that head losses in siphons can be significant, particularly in flat coastal areas, where the low terrain does not allow for surcharge and the available project corridor is narrow.

The size of the barrel or conduit can be determined initially based on the minimum required flow velocity. However, the barrel can be sized accurately only after he hydraulic losses are estimated. If the head loss under the design flow condition is excessive, increases in the size of the conduit should be considered.

Urban Stormwater Management Manual


Hydraulic Structures


300 mm


150 mm

i_. 300 mm


B —



Inlet Pipe

200 mm


I /.

2-R10 Rebar or Reinforce as Necessary

Figure 29.23 V-notch Level Spreader


Concrete Curb

Opening 280 mm

Grass Filter Strip Figure 29.24 Through-Curb Port


Urban Stormwater Management Manual

Hydraulic Structures


/ fir

/- Inlet Chamber            Outlet Chamber -^

^Manhole Frame and Cover

Removable Precast Concrete Slabs^


Inlet Chamber

Outlet Chamber

Concrete Encasement

Figure 29.25 Profile and Plan of Double-Barrel Siphon Vertical Legs (Engineering News, 1916)

400 mm Cl Bypass

1.5 m Circular Reinforced Concrete

(Storm) Pipes            —1.4 m x 1.2 m Reinforced Concrete Overflow

Old 1.6 m x 1.2 m Sewer

Dam Oulet Chamber

Cleanout Chamber

400 mm Cl (Dry Weather) Pipe (a) Sectional Plan

Dam Intake Chamber Cleanout Chamber

Old 1.6 m x 1.2 m Sewer

Cleanout Manhole


Cleanout Manhole

1.4 m x 1.2 m Reinforced Concrete Overflow

Subway 1400 mm Cl

B Structure

i-S m Circular Tw0 circular Reinforced Concrete Storm Pipes                       -------------------------------

Old 1.6 m x 1.2 m Sewer

z = z ID

New 600 mm Sewer

l°£s-----400 mm Cl

Section A-A

400 mm Cl (Dry Weather) Pipe

Section B-B

(b) Longitudinal Section

Figure 29.26 Profile and Plan of a Double-Barrel Siphon Sloping Legs (Engineering News, 1916)

Urban Stormwater Management Manual


Subject Index

Urban Stormwater Management Manual                                                                                                                                                           3 3