State of the art of the hydraulic design of stepped chute spillways M1, Chanson, University of Queensland, Australia Stepped spillways have recently become popular methods for handling flood releases, because of thei hydra inherent advantages associated with rapid energy dissipation. The ic design and several applications of stepped chutes are described in this, Paper, and the associated flow regimes and basic flow patterns are discussed. sin downstream of the spillway where a hydraulic jump is created to dissipate a large amount of flow energy; a high velocity water jet taking off from a flip bucket and impinging into a downstream plunge pool; or, the construc- tion of sieps on the spillway to assist in energy dissipation ‘Water flowing over a rough or stepped face ofa dam (for example, rockfill or a stepped channel) ‘an dissipate a major proportion ofits energy. For a stepped chute, the steps increase significantly the rate of energy dissipation taking place along the spillway face, and eliminate or reduce greatly the neod for a large onergy dissipator at the too of the spillway. Stepped spllways have been used for more than 2500 yeers (Chanson, 1994). Recently, new construction materials have in- creased the interest in stepped chutes. The stops increase significantly the rate of energy dissipa- tion on the spillway face and reduce the size of the required downstream energy dissipation basin. Stepped cascades are used also in water treatment. plants to enhance the air-wator transfer of atmos- Dherie gases (oxygen, nitrogen) and of volatile organic components. norgy dissipation over dam spillways usually achieved by: a standard stilling ba- 1. Applications In recent decades, stepped spillways have become ‘popular method for handling flood releases. Ex: amples of applications of stepped channels are presented in Table 1 and Figure 1. The construc- tion of stepped spillways is compatible with slipforming and RCC placement methods. Also, gabion stepped spillways are the most common type of spillways used for gabion dams. In Norway, steps were introduced in a tunnel spillway with free-surface flow to enhance the en- ‘ergy dissipation. This design was selected to allow detrainment upstream of a vertical shaft. and to prevent entrainment of air in the shatt. Stepped channels can also be used to inerease the discharge capacity. Russian engineers devel- ‘oped the concept of the overflow earth dam [Grinchuck et al,, 19774 Pravdivets and Bramley, 19899). The spillway consists of a revetment of precast concrete blocks laid on filter and erosion protection layer. The channel bed is very flexible ‘and allows for differential settlements: individual blocks do not need to be connected to adjacent blocks. A greater discharge capacity is achieved Hydropower & Dame duly 1994 ni ia (Table 1). The high degree of safety allows the use of such channels ag primary spillways [Pravdivets and Bramley, 1989"). In the UK, a stepped intake of a morning-glory type spillway was selected for the Ladybower reservoir, following model tests which showed that this arrangement provided a dischar than a smooth intake ig. 1. Beomples of stepped chutes sketch ofthe Kasserine dam, Tanisia (AD 100), showing detaile ofthe spillay; (ee also Photo below) BIS) The Chung dom, OK, whieh was completed in 1968 (Note the butiresees Denween each of the 10 lopped spillway) 28‘Table 1. Typical examples of stepped spillway ‘Name Ref. [Slope] Dam | Max. | Step [No. of] Type of steps Remarks Jneighe| diceh.|hoight) steps a | Ham | ae |b dog) [ad | aid | Concrete dame Clywedog,UK,1968 EN 60 72 28 0.176 Procast concrete ‘Buttress damm; 0 beams spillway made of ‘recast beams; W = 129m, DeMist Kraal,RS4, HD 88 30 50,119 Horizontalsteps © ROC dam, W= 195m 1986 Zasihosk,RSA,1886 HD 582 45 156 1 40_—_‘Hrizontalateps Monksvile,USA, S052 «965-93 (OL Horizontal steps 1987 Olivettos, France, BO «S81 86 «BG OG ~—AT_—Horizontalstepe 1987 co Upper Stilwater, HO -72& 61 116 O61 Horizontal steps RCC damm, W = 161m USA, 1987 59 M’Bali, CAR, 1990 BI 613 -24.5 16 «OB © 36 Horizontal steps. RCC dam, W= 60m PotitSaut,Guyana, DU 513 37 4 (06. Horizontal steps RCC dam oo Go ‘Barth overflow dams ‘Dnieper, 1976 = PR 8.75 60 0.405 12 Concrete block Full sate teste: V system; horizontal 23 m/s, W= 42m steps PR nas Stepped block system W= 12m ‘**Lukhovitsly dams, KR 3, Stepped block system ‘Three dams; W = 12, 1978, 1980, 1981 118, 33, 215m 15) 29) cTransbailal,1986 MI 140420 ‘Stepped block system Reservoir capacity = 5x08, W= 110 ‘Gobion dams Rictspruitoutfall, ST 18 2-4 Gabion stepe Cascade of three RSA weirs; W = 50 m “Masonry dams Gilbea, USA, 1926 GM 49° 18618 Inclined steps; 0=29 W=403.6m to5P Now Croton, USA, WE 84 Masonry steps Spillway failure in 1906 1955; W = 305m ‘Re-aeration cascades Calumet waterway, GA 152 aed Waterfalls ive artificial USA, 1991 «0st cascades Spillesay intake Ladybower, UK, 1939 EN 045 16 Masonry steps ‘Tulip intake of morning glory spillway ‘Tunnel epiltoay Stayord, Norway, = WP 113 O12 4 22 Pooled stops with _De-aeration system 1989 ‘Lanhigh wall ‘upstream of vertical shaft; L = 402 m Walined rock spiloay LaGrande2Canede PO 30 134. 16.14" 17.8 1 Smooth profiled stop; First step is concrete pool death = 85m" linad; W= 122m 91to 11 Unlined horizontal Second to last step is 122 stops unlined VI 15-219 11 501 4 © Smooth pooled steps; Smooth profiled 6 pool depths from 14 easeade system; four fem drops; W'= 60to 95m LY 12 12010" 2510 ¢ ~—Pooledstepe with Unlined rock. 35 10 mehigh walle Now: DI Dino etal (19891; D0 = Boouyge eal (188): DU = Duar tl (1983, BE = Engincerng (1905), EN = The Baginser [19851 GA = Guspartte 1991), GM = Gausmann an Maden [1925 HD = Hollingworth and Dray {Ig66), HO = Houston ane Richardson [19681 Kit = Krest'Yaninov and Frevewets (1085) LY ~ Lynne 00), Ml Miler eval [198TH 1.0.» Poat tal Li PR = Pravdivets und Bramley 188s; SO = Sorensen (845), St = Steptensan 1B), V1= Vital and Porey [1087 WE = Wega la907, Water Power (1990, * total discharge (10 mi) * = Former USSR." = Based on design. ou Hytropower Dams July 1994 ea Ea See eS Me2. Flow regimes ‘A stepped chute consists of an open channel with a series of drops in the invert. The total fall is divided into @ number of smaller falls. The flow over stepped spillways can be divided into two types of flow regime: nappe flow and skimming flow (Figure 2). oh tPTESetAL [1991 desribetwo types ofnappe low: I nappe flow with fully developed hydraulic jump for low discharge and small flow depth (Figure. 2a); and, Chappe flow with partially developed hydraulic Jump, also called partial nappe flow, (Figure 2b). The author (Chanson, 1994%] showed that a nape flow regime with fully developed hydrau- lic jump will occur for discharges smaller than a critical vale defined as: (2), -ams (4) WS etar t 9 where deis the critical flow depth, and t and J are the step height and length respectively. Nappe flow situations with fully developed hydraulie Jump oceur for de/lt < (d/h) char. Note that the correlation, Eq. (1), was obtained for: O2Sh/S6. Nappe flows are characterized by a succession of free-fall jets impinging on the next step and followed by a fully developed or partially devel- oped hydraulic jump (Figures 2a and 2b). Energy (dad hs In the skimming flow regime, the water flows down the stepped face as a steady stream, skim- ming over the steps, and is cushioned by the recirculating fluid trapped between them. The ex- ternal edges of the steps form a pscudo-battom, over which the flow passes. Beneath this, recirca: lating vortices develop and are maintained through the transmission of shear stress from the water flowing past the edge of the steps, For flat channels, the cavity of recireulating uid between the edges of adjacent steps is oblong Hydropower & Dams ly 1004 (Qe re Snweied Hywute Sait y i: / on Fig. 2. Flow regimes above a stepped chute: (a) Nope ow regime with fully developed hydraulic jump; @) ‘Noppe flow regime with partially developed hydraulic Jumps (e) Skimming flow regime with wake step inter- Teronces (Flat slopes: a < 27 degrees); (4) Skimmning flow ragime with wahe-wake interferences (Slopes : = 27 degrees: () Skimming flow regime with stable reire- lating cavity flow (Stsep slopes : > 27 degrees).Fig. 3. Geometry of large roughness 26 ‘and large stable recirculation vorticas cannot de- velop. The recircalating vortices do not fill the entire cavity between the edges, and the wake from one edge interferes with the next step [Baker, 1990']. The vortex generation and dissipation ‘processes associated with each wake are disturbed by the next step, and can interfere with those of the adjacent steps. The flow depth will control in part the vertical extent of the wake and the vortex interference region. ‘The main flow resistance pa- rameters become the distance between two ‘adjacent edges and the depth of flow (Chan- son,1994:). For very flat slopes, the flow pattern is characterized by tho impact of tho wake on the next step, a three-dimensional unstable recircula- tion in the wake, and some friction drag (skin friction) on the step downstream of the wake im- pact, This flow pattern is called a ‘wake-step interference’ rogime. For larger slopes, the tail of the wake starts interfering with the next wake, and the frietion drag component disappears. This pattern is called a ‘wake-wake interference’ ro- ime For steep slopes, a stable recirculation in the cavities between adjacent steps is observed as shown by photographs [Peyras et al. 1991'; Diez- Cascon et al., 1991"), The recirculating vortices are large two-dimensional and possibly three-di- mensional vortices. The energy dissipation and the flow resistance are functions of the enerey required to maintain thecireulation ofthese arge- seale vortices. This flow pattern is called a ‘recirculating cavity flow' regime. For flows past aecavity or above strip roughness, various researchers (Table 2) have observed & mechanism of stable recirculation for a range of the groove height-to-length ratio of the cavity (see Figure 3). Photographic evidence of stable flow recirculation can be found in Townes and Saber- sky [1968"] and Furuya et al. [1976"1. Most researchers have observed stable re-irculatory flow patterns for height-to-length ratios larger than 0.4. For a stepped spillway, the cavity height- to-length ratio equals: [cos a" sin al, and the value of 0.4 would correspond to a channel slope of a = 266°. The results summarized in Table 2 would imply that stable recirculation occurs for slopes larger than 26.6" on a stepped chute, The author [Chanson, 1994'] showed clearly a differ- tent flow resistance behaviour for channel slopes larger than 27°, and this is consistent with the findings of the other researchers (Table 2). For slopes of around 27°, the flow pattern is charac- terized by the interference of the wake tail with the next wake. ‘The flow characteristics of the skimming flow regime were extensively analysed by the author [Chanson 1993°; 19947). Skimming flows are char- acterized by large friction losses and a strong air entrainment process. In the case of large dams, it thas been shown that they dissipate more energy than nappe flows (Chanson, 194°]. ‘Table 2, Observations of stable recirculatory flow motion Reference Condition for stable reeiroulation Remarks Water flow experiments "Townes and Sebersky {1965} Knight end MacDonald (1973] Rectangular wooden bars Wind tunnel experiments ‘Maull and East (1968), Kistler and Tan (19671 Flow past a two-dimensional rectangular evity Flow past a two-dimensional rectangular cavity, ‘Two-dimensional circu roughness * trip beteon the water ges ofthe dela rods. Not: (La) = grove or ceviy length manared parallel toe Mow direction Pig Oh ™ theo vols" = (EI menue Hadropower & Dams July 1994 Seat3. Air-water gas transfer 5.1 Introduction ‘One of the most important water quality parame- ters in rivers and streams is dissolved oxygen (DO). The oxygen concentration is a prime indica- tor of the quality of the water. Dams and weirs across a river affect the air-water gas transfer dynamics. The flow conditions above a stepped chute are characterized by a high level of turbu- lence, and large quantities of air are entrained (Figure 4). Turbulent mixing and air entrainment contribute to an enhancement of the air-water gas transfer along stepped chutes, as compared with smooth chutes. ‘The gas transfer must be taken into account for the re-oxygenation of polluted streams and rivers, but also to explain the high fish mortality downstream of large hydraulic structures caused by nitrogen supersaturation. 3.2 Applications In rivers, artificial stepped cascades and weirs can be introduced to enhance the dissolved oxygen content of polluted or eutrophic streams, [for ex- ample, Avery and Novak, 1978", Nakasone, 1987], In the USA, the Tennessee Valley Author ity (IVA) designed’ and built in-stream cascades and drop structures to re-oxygenate rivers [Hauser et al., 199244), For large dams, the downstream nitrogen con- tent is another important parameter. Nitrogen supersaturation may increase the mortality of some fish species (Smith, 1973"). In the treatment of drinking wator, cascade aeration is used to remove gases (volatile organic compounds or VOCs), and to eliminate or reduce offensive taste and odour. Inclined corrugated channels or drop structures are effective means for water treatment [Boyden et al., 1990", Corsi et al., 1992"), The combination of flow aeration and high flow turbulence enhances greatly the mass transfer of VOCs such as chlorine. When ower is not required, this process can be success- ful, since the operation requires little main- tenance. 5.3 Nappe flow regime Ina nappe flow regime, the air-water gas transfer at each step results from the gas transfer at the plunge point and at the downsiream hydraulic Jump (Figure 4a). The aeration efficiency at one individual step equals: 5-5" +8 (1-24) (3) where Dis the aeration efficiency of the plunging jet flow, Eis the aeration efficiency of the hy- ‘raulic jump and the aeration efficiency is defined where Cusis the upstream dissolved gas content, Cos is the downstream dissolved gas content and Cris the gas saturation concentration in water. ForaseriesofNripsteps, the aeration efficiency of the complete cascade can be expressed in terms of the aeration efficiency of each step: ydopowee & Dams Joly 1994 ae ™< we - ‘eapny (4) where Fis the aeration efficiency of the i-th Step, estimated using Eq. (3). For a stepped chute with Nui steps of equal characteristics, the aeration efficiency of the chute becomes -(-5)"" 6) ‘The total air-water gas transfer results from the successive aeration of the flow at each step. At each individual step, the flow is aerated by a com- bination of air bubble entrainment at the intersection of the falling nappe with the receiving pool of water on the step, and air entrainment at the following hydraulic jump. ‘The mechanisms of air entrainment by plun, ing jet and by hydraulic jump are complex. Although experimental investigations are under ‘way at several laboratories, no prediction method is yet available and the author believes that eccu- rate prediction methods will not be available for some years yet For design purposes, a crude estimate of the dissolved oxygen transfer can be obtained by ‘feeding’ empirical correlations [Avery and No- vak,1978"} into Eqs. (4) or (5). The results can be double-checked with overall correlations. Such approximations might provide an order of magni- tude of the downstream DO content if the flow conditions are within the range of validity of the empirical correlations; but they will become very Inaosarate outside the range of validity ofthe cor relations. 3.4 Skimming flow regime On a stepped spillway with skimming flow, the entraining region follows a region where the flow over the spillway is smooth and glassy. Next to the ig 4 Air entrain. ment on stepped spillways:() Nappe fiw regime; 6) Skin ming flow regime. ea reece See eee CeCeboundary, however, turbulence is generated and the boundary layer grows until its outer edge reaches the surface (Figure 4b]. When the outer edge of tho boundary layer reaches the free sur- face, the turbulence can initiate natural free surface aeration [Chanson, 1993°]. Downstream of the inception point of air entrainment, the flow rapidly becomes aerated, as shown by the photo. graphs [Krest'yaninov and Pravdivets, 1986" Diez-Cascon et al,, 1991"; Pyras et al, 1991"), Using some similarities between free-surface aerated flows and stepped chute flows with skim- ming flow regime, the air-water gas transfer of skimming flows on stepped chutes can be esti- mated using the same method as Chanson (1993). The complete method is presented else- where (Chanson, 1994] and the main findings are presented below. ‘The author carried outa series of oxygen trans- fer calculations in self-aerated flows on stepped chutes, assuming an ungated spillway, zero salin- ity, constant channel slopesranging from 15 to60", channel lengths of between 20 and 250 m, dimen: sionless discharges dd/t from 0.8 to 21, a friction factor f = 1.0 and temperatures between 7 and 30°C, ‘The analysis of the oxygen and nitrogen trans fer computations indicates that the free-surface aeration efficioney is independent ofthe initial gas content. Furthermore, the aeration efficiency (for ‘oxygen and nitrogen) can be correlated by: Go (), 6 where (gu) is the discharge for which the growing boundary layer reaches the free surface at the spillway end and no self-aeration oveurs, and the exponent A is a function of the diesolved gas, the temperature and the spillway slope. For the range of the computations, the exponent A ranges be- tween 8 and 9, but no simple correlation has been obtained between A and the flow characteristics ‘The characteristic discharge (g,). can be deduced from Chanson (1994"]: (Ga), =01277-(Litay) (sina)? (1*c050)°"™ =) where Lipitsyis the spillway length and a is the spillway slope Froe-surface-aeration calculations indicate that the aeration decreases with an increase in discharge. Furthermore, the aeration efficiency resulting from free-surface aeration increases with increasing channel slopes from 15 to 45", and isalmost the same between 45 and 60°. Fora given discharge, an increase in channel slope eauses an. increase in the amountof air entrained, and hence fan increase of the interface area, Tie mean veloc. ity also increases with the channel slope, and hence the residence time daereases with increas: ing slopes. For flat slopes, the quantity of air entrained is not large enough to obtain optimum aeration, For steep channels, the mean flow veloc. ity might become too large and the residence time too short. For a stepped chute with skimming flow, the computations suggest that: S the air-water gas transfer caused by self-aera- tion increases with decreasing water discharges for a constant channel slope; and, the aeration efficiency is at maximum for chan- nel slopes ranging between 45 and 60", for a given dischargo, A decrease in water discharge reduces the length of unaerated flow. Hence the gas transfor caused by free-surface aeration takes place over a longer distance. However, the designer needs to Keep in mind that the discharge must be lange ‘enough to satisfy the conditions of the skimming flow rogime, Eq. (2). 4, Applications: design of stepped chutes 4.1 Introduction ‘The design of a hydraulic structure is a Jong process which involves many external factor economies, environment, geology, hydrat 5, hy. drology, politics, sociology, and so on. For the design of a channel, for example a chute or spill- way, the first questions to ask are ‘© What is the purpose (the duty) of the structure? (For example, energy dissipation, water ieeat, ‘ment, emergency spillway, flood release); and, © What are the constraints? (For example, geome try, hydrology, material availabiligy, von. ography, geology, polities.) When the decision to select a stepped channel hhas been made, all the parameters must be aken into account to decide on the preferential type of Alow regime (nappe flow regime, or skimming flow regime) for the design discharges and for unusual (emergency) flow conditions (in drought or fool periods). It should be emphasized that flow conditions near the transition between the nappe and skim- ting flow regime must be avoided Transitory fluctuations between the nappe and skimming flow regime can induce unfavourable or even dane gerous flow behaviour and unnecessary structural vibrations. 4.2 Concrete stepped spillways 422.1 Introduction Stepped spllways can be used to assist in energy dissipation and to help reduee the size (and cost) of the downstream stilling basin, Comparisons of the energy dissipation with nappe flow regime and skimming flow regime (Chanson, 1994) showed that: © for long channels where the uniform flow con- ditions are reached, the greatest energy dissipation is obtained with a skimming flow regime; and, © for a short stepped chute, a nape flow regime enables more energy dissipation to be achieved than a skimming flow regime, 4.2.2 Design procedure A typical design procedure includes the following steps: 1. Assessment of the design flood, Hyropower Se Dams duly 1294 (ee2. Assessment of the spillway geometry: slope, height and width. The spillway slope might be a. design variable in some cases. A steep slope can. minimize the volume of the structure, but greater energy dissipation is achieved with Mat slopes. 3. Selection of the optimum step height. The step height is chosen as a function of the discharge and the channel slope to obtain the selected flow regime (nape or skimming flow) andehe appro- priate level of energy dissipation. 4, Selection of the step length (or channel slope). Ifthe channel slope is not fixed by the topography or the method of construction, the step height and the channel slope must be selected to provide the optimum flow conditions and/or energy dissipa- ton: 5. Calculation of the hydraulic flow charac- teristics. For the selected step geometry, the flow. depth, velocity, the amount of air entrainment and the energy dissipation can be calculated (Chanson, 1993", In particular, it is important to take into account the effects of free-surface aeration forthe design of the chute sidewalls. 6. Caleulation of the downstream dissolved gas contents (oxygen and nitrogen). For the climatic conditions of the site, and the chemical properties of the water, the aeration efficiency of the struc- ture must be estimated (see Section 3). The most. usual parameters are the dissolved oxygen and nitrogen gas contents. Prediction of the down- stream dissolved gas contents must be made for & wide range of discharges and atmospherie condi- tions Spillway erest Various shapes of spillway crest have beon used for the crest of stepped spillways. For skimming flows, model observations have shown the risks of de- flecting jets of water at the first steps if these stops ‘are too high. Some researchers (Sorensen, 1985"% Beitz and Lawless, 1992; Bindo et al., 1993°4) suggested introducing. few smaller stepsnear the crest to eliminate the risk of deflections. ‘The design of the first steps is still empirical, and is deduced from model studies in most cases. Nappe aeration For @ nappe flow regime, the vent of the free-falling nappe is essential, to avoid unsteady fluctuations resulting from nape oscillations. In general, the most convenient method for ventilat- ing the’ nappe is the use of sidewall deflectors, ‘which provide space between the sidewalls and the deflected nappe. The design of these deflectors is similar to the design of aeration devices (Wood, 1991"). It is also possible to aerate the nappe by small ducts feeding air ftom beyond the sidewalls to near the top of each drop. If these techniques are not feasible, splitters may be used to break the sheet of free-falling water, allowing space for aera- tion and preventing nappe fluctuations. 4.2.3 Roller compacted concrete (RCC) structures Since the late 1970s, RCC has become a popular material for the construction of gravity dams. The main advantages of RCC gravity dams are their cost effectiveness, reduced foundation surface Hydropower & Dams duty 1994 area compared with earthfill or rockfil structures of the same height, and short construction time. ‘The low cost of RCC gravity dams results from: ‘the small material volumes (vertical upstream face and steep downstream face); Slower cost per unit volume compared with con- ventional concrete; 8 simpleconstruction techniques (small amount of formwork, placement using standard earthwork techniques}; and, ‘reduced cost of auxiliary structures (intakes at- tached to the vertical face, stepped spillway on ‘the downstream face), in comparison with earth- fill embankments. With a roller compacted concrete dam, itis usual to shape the steps with a protective layer of medium to high-resistance concrete (Figure 5). ‘The joint between the RCC and the conventional concrete must be made with great eare. Indeed model tests (Frizzell and Mefiord, 1991] have shown that the vertical face of the steps can be subjected to sub-atmospheric pressures. With a stepped spillway on RCC dams, itis important to incorporate into the design a drainage system bo- hind the erosion-resistant layer. Por small discharges and a temperate climate, the steps might be simply formed RCC. Hansen [1986] reported eases of flow over a stepped RCC face at the Brownwood Country Club dam. There the spillway was subjected to flow depths of 0.05 m in one case and 0.3 m in another, without damage, despite the absence ofa protection layer. In colder climates, unprotected ROC surfaces can experi- ence some fracturing or chipping during freeze-and-thaw cycles, A conventional concrete overlay can be applied at the end or after the Gompletion of the construction, to protect the 4.5 Gabion stepped spillways Gabions are extensively used for earth retaining structures and for hydraulic structures (weirs and channel linings). Their advantages are stability, low cost, flexibility and porosity. ‘The design of gabion stepped spillways is lim- ited by their resistance to abrasion and destruction, and their stability. Various re- searchers have proposed criteria to prevent the destruction of gabion spillways. Stephenson [1979 discussed the criterion for structural sta- Dility of both stacked and aligned gabions. For stacked (stepped) gabions, Peyras et al. [1991"] suggested the combination of a concrote slab (6 to 10 cm thiek) on the step surfaces to enhance the abrasion resistance. Inclined (upward) gabion- stepped spillways can also be used. Greater energy Fig. 5, Details ofthe construction of steps st RCC dams 2cd Fig, 6, Details of concrete blocks on a earth dam spillway: (a) lat concrete blocks (rectangular cross-section} () flat concrete blocks (iropezoidal cross-setion);(e) wedge-shaped concrete blocks dissipation is achieved, but their construction re- quires greater care (Peyras et al., 1992". In any case, the stone sizeof the gabion filling must be at least 1.5 times the mesh size, Gabion dams and gabion structures might be subjected to another risk of destruction. In some countries (for example in parts of Africa or South- America), the inhabitants might use the gabion ‘wires for other purposes (for example fences) or the gabion filings as construction material. In such places, gabions should not be selected as con- struction material tis interesting to note that the discharge sta- Dility criteria for gabion spillways are of the same ‘order of magnitude as the recommendations for ‘unprotected rockilled channels. ‘The dimensions of the gabion and the design discharge will determine the type of flow regime. ‘The height of the step equals the height of the gabion, Usually the step length equals 1/2 to 1/4 of the gabion length. With these dimensions, Eqs. (and (2) will indicate the type of flow regime for the design discharge. If the flow conditions are close to the transition between nape and skim- ming flow regime, the dimensions of the gabion or the width of the chute must be modified: either to change the step height h or to alter the discharge per unit width gv. For a structure made solely of gubions, the choice of a steep slope with a skimming flow re- gime would reduce the number of gabions and the cost of the structure. For an earthfill structure protected by gabions, a flat slope and a nappe flow regime may be more appropriate for the stability requirement of the embankment. 4.4 Earth dam spillway with precast concrete blocks 44.1 Introduction In recent years, several old embankment dams have bean identified as being unable to pass the design flows without failure caused by overtop- ping. Additional discharge capacity can be obiained by designing a spillway made of concrete blocks on the downstream slope of the embank- ment. Russian engineers were among the first to pro- pose a stepped concrete protection on the downstream face ofa dam to pass flood discharges. ‘The choice ofa stepped structure allows the use of individual blocks interlocked with the next ele- ments (Pravdivets and Bramley, 19895), and assists in energy dissipation. For new dams, a stepped spillway made of concrete blocks may be considered as the primary flood release structure of the embankment [Miller etal, 1987) 442 Protective layer ‘The most important aspect of an earth dam spill- way with a stepped overlay is that the dam material must remain stable. Seepage may occur ina saturated embankment and in such a ease, the resulting uplift pressures could damage or dé the stepped channel and the dam. ‘Table 3, Design and stability of concrete blocks for embankment dam spillways Reference Max Slope Block shape Block Comments isch. thickness (ats) deg) (a) Grinchuk otal (1977) 0 81 Wedgeshape reinforced 0.7 Dleper hydro plant; (Wa22 mvs) concrete blocks downward inclined etape Pravdivels and Bramley 2027 Tto15—Wedge-shaped block (0.15 0045. Downward inclined 1983) construction steps Baker (1980) 32.7 Flat blocks (trapezoidal ‘Model tests; downward section); downvtard sloping 218 — Wedge-shaped blocks; downward sloping inclined steps No fur reported for V<6.51 me ‘Reinforced grass water wey Hewlett etal. [1987] (V<4to——21.8_—_—Interlocking non-tied Tm) canerete blocks Wet 218 Bra) Hytropower & Dams Joly 199¢ ennAdequate drainage is thus essential, For a typi- ‘al design, the blocks lic on a filter and erosion protection layer (Figure 6). The function of this layer isto filter the seepage flow out ofthe subsoil and to protect the subsoil layer from erosion by flow in the drainage layer. The protection layer also reduces or eliminates uplift pressures acting on the concrete blocks. Frizell (1992! showed that suction ofthe fuid from beneath the concrete steps can be produced by the pressure differential created by the high velocity flow over the step offset area (that is, the vertical face of the step). Drains placed in areas of sub-atmospherie pressure will function to relieve uplift pressures, In any case, the location of the embankment drains musi be appropriately se- lected to avoid reverse flow in the drains and dynamic pressures associated with hydraulic jumps (in the case of a nape flow regime) Grinchu et al. [19774] recommended that the to- tal area of the drainage holes should be 10 to 15 per cent of the block surface arsa. Noto that Frizell [1992" studied flows past horizontal steps and steps sloping downwards at 10 and 15°. Her model tests suggested that the negative pressure (on the vertical lace ofthe step) ‘nereased with the downward slope of the stops ‘The seepage flow in the embankment dam must be predicted accurately so that appropriate provi- sion can be mado for drainage and evacuation of seepage flow though the blocks. Note thal the seepage flow may be influenced by the infiltration {nto the downstream slope caused by the spillway flow, in addition to the flow through the embank ment. 4.4.3 Block geometry and configuration Russian engineers developed considerable exper- tise in the design of concrete wedge blocks Pravdivets and Bramley [1989] described several configurations (Figure 6 and Table 3). The blocks ‘are normally laid parallel to the slope on the top of the protective layer. Individual blocks do not heed to be connected to adjacent blocks. For large Gischarges, the blocks may be tied to adjacent blocks and would normally be of reinforeed con- crete, ‘An interesting option is the use of reinforced grass waterways. Hewlett et al. [1987 tested ‘grass waterways with a variety of reinforcement methods (textiles, nylon, and conerate blocks): grass waterways, reinforced with concrete blocks, could sustain velocities of up to 4 to 8 mis without damage (Table 3). Several grass species can be used, for example, bent, eouh, fescue, kikuy, meadow grass, and perennial ryegrass, 444 Design considerations Pravdivets and Bramley [1989] recommended strongly thatthe alignment ofthe spillway should be straight from the erest to the te. Any curva ture of the spillway in plan, oF change in cross-section, will equse an uneven distribution of flows within the spillway which, in general, should be avoided". Usually the channel sidewalls are Nat inclined slopes. The slopes ofthe sidewalls ean be designed as inclined stepped surfaces (in the flow direction) and may use the same conerete block system as Hydropower & Dams. July 1994 the main channel. Pravdivets and Bramley sug- gested typical side slopes of about 1V:3H (thet is, 18°. 5. Conclusion Stepped spillways are selected to contribute to the stability of the dam, for their simplicity of shape, and to dissipate the energy of flow. The steps in- crease significantly the rate of energy dissipation taking place on the spillway face, and reduce the size of the required downstream energy dissipa- tion basin. On a stepped channel, five different flow regimes can occur: nappe flow regimes with fully developed and partially developed hydraulic jump, and skimming flow regimes with wake-step interferences, wake-wake interferences and stable recirculating cavity flows. ‘The practical cases which have been discussed here demonstrate the state of the art of hydraulic design of stepped spillways, including RCC dam spillways, gabion stepped spillways and earth dam spillways with precast concrete blocks. ° References 1. Chanson, H., "Hydraulics of skimming fows over stepped channels and splays." Journal of Hydraulic ‘Research, IAHR, Vol 32, No. 3; 1994, 2. Grinchuk, AS., Pravdivets, ¥.P.,and Shekhtman, NV "Tea of earth slope revelmenis permitting Now of water at large specific discharge’, Gtdroiekhnicheskoe Stroiel'stoo No. 4, (in Russian); 1977. (Translated in Hydrotechnieal onstruction, ‘Plenum Publishing: 1978). 8, Pravdivets, YP. and Bramley, MLE. “Stepped protection blocks fordam spllways" Water Pouser & Dam onstruction; July 1988 4. Peyras,L, Royet, P., and Degoutte, G,"Ezoulement et dissipation cur lee déversoirs en gradine de gabions” (Plows and dissipation of energy on gabion weirs) La Hlouille Blanche; No.1, 1991, 5. Chanson, I, "Hydraulics of nappe flow regime above stopped chutes and spllways" Australian Civil Engi- neering Transactions, Vol. CB36, No. 1; January 1994 6 Chanson, HL, "Comparison of energy dissipation be- ‘tween nappe and skimming flow regimes on st=pped chutes" Journal of Hydraulic Research, IAHR, Vol. 32, No, 2 1994 7. Essery, LTS, and Horner, MW., "The hydraulic ‘design of stepped spillways" CIRLA Report No.3, 2nd edition, London, UK; January 1976, 8, Chanson, H, "Stepped spillway flows and air entrain. ‘mont * Canadian Journal of Civil Engineering, Vo. 20, No. 3;June, 1993. 9, Baker, Ry "Procast conerete blocks for high velocity ‘low applications" Journal of IWEM, Vol. 4, December 1Dier-Cascon, Ju, Blanco, J-L., Revilla, J, and Garcia, R., “Studies on the hydraulic behaviour of stepped spillways” Weter Power & Dam Construction: September 1991 11.Townes, H.W., and Sabersky, FLH., “Experiments on ‘te flow overs rough surface” International Journal of Heat and Mass Transfer, Vol. 9; 1968 12Furuya, ¥., Miyata, M., and Fujita, H., "Turbulent ‘boundary layer and How resistance on plates roughened by wires" Journal of Pluids Engineering, Transactions, ASME, Vol. 98; 1976, I.Avery, ST, and Novak, P., "Oxygen transfer at hydraulic strueturee" Journal of Hydraulics Division, ‘ASCE, Vol. 104, No. HII; 1978,