Embed Size (px)
Citation preview
SPILLWAYS FOR DAMS
Commission I nternatioш1le des G rands Barrages 151, bd Haussmann, 75008 Paris - Пl. : 47 64 67 33 - Пlех : 641320 F ( I CO LD)
NOTICE – DISCLAIMER :
The information, analyses and conclusions referred to herein are the soleresponsibility of the author(s) thereof.
The information, analyses and conclusions in this document have no legalforce and must not be considered as substituting for legally-enforceableofficial regulations. They are intended for the use of experiencedprofessionals who are alone equipped to judge their pertinence andapplicability and to apply accurately the recommendations to any particularcase.
This document has been drafted with the greatest care but, in view of thepace of change in science and technology, we cannot guarantee that itcovers all aspects of the topics discussed.
We decline all responsibility whatsoever for how the information herein isinterpreted and used and will accept no liability for any loss or damagearising therefrom.
Do not read on unless you accept this disclaimer without reservation.
TABLE OF CONTENTS
LIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . 7
FOREWORD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1. INTRODUCТION . . . . . . . . . . . . " .... """ ............ " ...... " .. "" .. " ...... " ........ " ........ " .... "" .... .. "".. 11
2. SELECTION OF SPILLWAY ТУРЕ . . . . "" .... "" .. "" .......... " ........ "" ......... ".""""""" 13
2.1. Spillway types and classification "" .. . . . " ... " .... "" .. . . . . . . """ ...... "" .... " .......... "...... 13
2.2. Factors in selection of spillway type . . . . . . " .. " ................ """""""""""""" .... """ 15
2.2.1. Flood prediction """"""""""""""""" ...... "".""."" .......... " ...... " ....... "."" 15 2.2.2. Seismicity of project area and reliaЬility of operation """"""""". . . . . . 17 2.2.3. Duration and amount o f discharge . . . . " ............ "".""""""."""""""""" 17 2.2.4. Site topography and geology """""""."."""" ...... "" .......... " .... ""............ 19 2.2.5. Dam type . . . . . . . . . . . . . . . . . . """."."""" .. """"""" .............. "................................ 19 2.2.6. Operational arrangements . . . . . . . . . . . . . . . . " .. " .... """"""".".""""""""""."."" 21
3. SURFACE SPILLWAYS """"""""""" .................... """""""""""."""."""".""""". 23
3.1. Components "" .... """"""" .. "" .................... " .... "." ..... " .... " .... """" .. "" .. """""".. 23
3.2. Gated, ungated and comЬination surface spillways . . . "" ... """"""""""""".". 23
3.2. 1. Ungated spillways " .""" ..................................... ".""".""".".""."."""""" 23 3.2.2. Gated spillways " .................... """"""""""""""".".""""""""" .. """"...... 25 3.2.3. ComЬination spillways . . . . . . . . . . . . "" .... " .. """""" .. "" .. """"""""" .. """"" .. " 25
3.3. lnlet geometry . . . . . . """ .. " .... "."."" .. " ........ """""" .......... "" ... " . ........ """"""......... 25
3.4. Gate types, arrangement and selection """""""" ........... "" ............... " .......... "" 27
3.5. Chutes .. " ............ """"""""""" ............ " .......... """"""" .. """""."." .. ".................. 39
3.6. Terminal structures and energy dissipation . . .. . . "" ........ "" .. """"""""""""""" 5 1
3.6.1. PreamЫe """"". ". """""." ." .................... "" .... ".".""""""""""""."""""". 5 1 3.6.2. Stilling basins : accidents, hydrodynamic proЫems, uplift pressures,
vibration, cavitation, abrasion, maintenance """." ... " ................ " .. "".". 55 3.6.3. Ski jumps and nappes, scour holes, erosion, aprons, hydrodynamic
proЫems . . . . " .................... "."".""""""""" .............. " ........ "" .. "" .. """"".". 69 3.6.4. Model tests """" .. """"" ............................ """ .. """"" .. """""""""............ 9 1
3.7. Protection against ice . . . . . . . . . . . . . . . . . . . . "." ... """""."""""" ... " .. " .............. " ............ " 93
4. ORIFICE SPILLWAYS "" .................... """"""" .. " .......... " ...... ""."""""""."".......... 95
4.1. Components, leading characteristics, objectives . . . . . . "" .. " ... " .. ""."."."""""".. 95
4.2. General characteristics of sluices . . . . . . " ........... """"".""" ".""" .................. "....... 105
4.2.1. Large bottom outlets . " ........ "." """ """ "."." .. " .............. "" .". """""""""" 105 4.2.2. Compensation outlets "" ............... "."""""""" .................... "" .. """ .. """" 107
4.3. Gates and valves """ .... " ........... """""".""." ... " .... " ....... """""""""."." ......... "" 107
4.3.1. Bottom outlet gates """".". "" ............... " .. """""""."""" ..... " ................. " 107
3
4.3.2. Maintenance gates for high-capacity bottom outlets . . . . . . . . . ... . . . . . . . . . ... . . . . 1 15 4.3.3. Gates and valves for fine control .. . .. . . . . . . . .... . . . . . . . . . ... . . . . . . . . .. . . . . . . . . .. . . 1 15
4.4. Cavitation, steel lining, maintenance (see also Chap. 5 and 6) .................... 117
4.5. Protection against floating debris . . . . . . . . .. . . . . . ......................................... 119
4.6. Protection against siltation ... ............................................................................ . 119
4. 7. Protection against vorticies .. .. .. .. .. ........ .. ............ .. .. .. ........ .. .. .......... ...... .. .. ........ .. .. 121
4.8. Terminal structures and energy dissipation (see also Section 3.6) ................ 121
5. PARТICULAR PROBLEMS WIТH HIGH-YELOCIТY FLOWS
5.1. Cavitation : General considerations, typical examples ................................. .
5.2. Control of cavitation damage .......... ........................................ ......................... .
5.2.1. Surface finish, practical limitations, concrete ageing ......................... .
5.2.2. Surface treatment, linings and coatings ................................................. .
5.2.3. Aeration ................................ . ........................................ .................. .
5.3. Abrasion in spillways and bottom outlets - Special surfaces ................. .
5.3.1. General " .. """"""""" .... . . . . . . . . . . . . .. . . . . ... . .. .... . . . . . . . . . . . . . . . . . . . . . . . . .. .. . . . . . . .
5.3.2. Abrasion from bed load .......... ....................................... ................. .
5.3.3. Abrasion from suspended load ................ .............................................. .
5.3.4. Surface protection ........................... " . . . . . . ... . . . . . . . . " .. . . . . . . . . . . . ... . . . . . . . . . . .... . . . . . . . .
5.4. Nitrogen release in spillway and outlet tailwaters
123
123
127
127 127 133
145 145
145 147 151
153
6. MAINTENANCE AND REPAIR . . . . . . . . . . .. . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . ... . . . . ".. 157
6.1. Maintenance . . . . . . .. . ......................................................... ................................... 157
6.2. Repairs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
APPENDICES ... ........................................................................................................ .
1. References .... ................... ........................................... ....................................... .
157
162
162
2. Countries which have answered the enquiry оп spillways . . . . . . . . .. . . . . . ... . . . . . . . . . . . . . . . . . 172
5
LIST OF FIGURES
Fig. 1. - Spillway of Karun dam (lran).
Fig. 2. - Ku-Kuan dam spillway (Ta'iwan).
Fig. 3. - Morrow Point dam (USA). Spillway orifices.
Fig. 4. - Itaipu dam (Brazil). Spillway and spillway gates.
Fig. 5. - Youssef Ben Tachfine (Morocco). Surface spillway.
Fig. 6. - La Grande 2 dam (Canada). Spillway.
Fig. 7. - Tarbela dam (Pakistan). Stilling basins of Tunnels 3 and 4. Fig. 8. - Malpassao dam (Mexico). Plan view and longitudinal profile of spillway.
Fig. 9. - Sidi Mohamed Ben Aouda dam (Algeria). Spillway.
Fig. 10. - Hendrik Werwoerd and Р. К. Le Roux dams (South Africa). Splitters of surface spillways.
Fig. 1 1. - Tarbela dam service spillway (Pakistan).
Fig. 12. - Permanent additional pressure over the apron surface below а free-falling jet.
Fig. 13. - Kariba dam (Zimbabwe). Section through orifice spillway.
Fig. 14. - Cabora Bassa dam (Mozambique). Mid-depth orifice spillway, longitudinal section.
Fig. 15. - Sainte-Croix dam (France). Bottom spillway, longitudinal section.
Fig. 16. - Kashm el Girba (Sudan), Jupia (Brazil) and Roseires (Sudan) buttress dams. Typical sections of bottom spillways.
Fig. 17. - Aldeadavila dam (Spain). Section through bottom outlets.
Fig. 18. - Р. К. Le Roux dam (South Africa). Bottom outlet and details of control slab gate.
Fig. 19. - Spillway tunnel of Yel\owtail (USA). General layout and details of aerator.
Fig. 20. - Ust llim dam spillway (USSR). General layout and details of aerators.
Fig. 21. - Foz do Areia dam spillway (Brazil). General layout and details of aerators.
Fig. 22. - Tarbela dam irrigation tunnel No. 3 (Pakistan). Aerator at top of stilling basin chute. General layout and details of air trough.
7
M EM B E RS O F Т Н Е СОМ М IТТ Е Е O N H YD RA U LICS FOR DAM S
М . CA R L I E R ( Fraпce/ Fraпce), Presideпt - Chairmaп К. B EL BAC H I R (Algerie/ Algeria), Membre - Member Е. C U R I E L (Veпezuela/Veпezuela), Membre - M ember F. G . D E FAZIO ( Etats-Uпis/USA), Membre - M ember J. K NA U SS ( Rep. Fed. d'Allemagпe/ Fed. Rep. of Germaпy), M embre - M ember G. MA R I N I ER (Сапаdа/Сапаdа), Membre - M ember М. M E N DI LUC E ( Espagпe/Spaiп), Membre - M ember N. PI NTO (Bresil/ Brazi l), Membre - M ember А. A LVA R EZ RI B E I RO ( Portugal/Portugal), M embre- Member С . Р. ROB E RTS (Afrique du Sud/South Africa), M embre - M ember У. S E M E N KOV ( U RSS/ U S S R), M embre - Member J . Н. SO N U ( Rep. de Coree/Rep. of K orea), M embre - M ember J . TEJA DA (Colombie/Colombia), M embre - M ember М. V E RCO N ( Yo ugoslavie/Yugoslavia), Membre - M ember
FOREWORD
This Bulletiп has Ьееп prepared оп behalf of the Freпch N atioпal Committee for the Committee оп Hydraulics for Dams Ьу G. Post апd L . Chervier (Соупе et Bellier - Fгапсе) for the chapters 1, 2, 3, 4, 6 апd аппех 1, апd Ьу С. Blaпchet апd G. Johпsoп (Sogreah - Fгапсе) for Chapter 5.
The draft was discussed апd fiпalized Ьу the Committee members at the aпual Committee meetiпgs of 1984, 1985 апd 1986. The fiпal versioп was approved Ьу the Executive M eetiпg iп 1986.
The origiпal text iп Freпch has Ьееп revised Ьу the C hairmaп М . Carlier (Fгапсе) . The Eпglish traпslatioп has Ьееп dопе Ьу R. Chadwick апd checked Ьу Е. J . Beck апd F. G. DeFazio ( U SA) апd С. Р. Roberts ( South Africa).
The coordiпatioп was Ьу М . Carlier.
9
1. I NTRODUCTION
А consideraЬ\e and varied body of experience has been built u p over recent decades in the design and construction of outlet works for large dams, and in 1 978 , the C ommittee on H ydraulics for Dams of ICOLD sent out questionnaires to review the salient features on th is subject. This report is based on the information tendered, a\though unfortunately very inadequate in terms of the number of replies received and the detai ls contained therein, and papers submitted to recent I C O L D Congresses.
The basic design of spil lways involves а variety of complex proЬ\ems and, without pretending to Ье а comprehensive state-of-the-art review, this report does summarize the main considerations arising from modern dam projects. In the questionnaire the scope was quite l imited since some aspects were given greater weigh t ; the following points, for example, were excluded since most of them were already dealt with in previous puЬ\ications and are not therefore discussed hereafter : selection of design flood, dealing with water during dam construction (see ICOLD Bul letin 48), detai led structural and stability design (see I C O L D Bul letin 27) , operation (see ICOLD Bul letin 49), overflow dams and weirs, combined spil lway/hydro powerstation structures, breaching dykes, energy dissipation and scale model studies. The report does however contain some remarks on energy dissipation and scale models insofar as they are intimately related to the determination of spi l lway design and size.
Chapter 2 begins with а brief summary of spillway types, and then goes on to examine the many factors involved in choosing between them, ie discharge capacity, head, number of weeks' or months' spill ing each year, dam type, topography, geology, flood detention capacity in the reservoir, and operating arrangements. The diversity in the types of spi l lway and gate plant used arises from the number and variety of such factors.
For simpl icity, the report fol lows the conventional distinction between surface spi l lways and submerged spillways. А separate chapter is devoted to each ш an attempt to identify their specific proЬ\ems and critica\ features (Chap. 3-4) .
Two other chapters (Chap. 5-6) are devoted to proЬ\ems arising out of high ve\ocity flow ( cavitation damage, abrasion, nitrogen release and precautionary measures), and spil lway maintenance and repair.
Wherever possiЫe, each chapter has examples of typica\ dams currently in operation throughout the world where sufficient information is availaЫe from the enquiry or other sources.
References appear in Appendix 1 and countries which have answered to the questionnaire in Appendix 2.
Names of the dams incl uded in this bul letin are fol lowed, between brackets, Ьу the name of the country, province or territory, where the dam is \ocated.
1 1
2. SELECTION OF SPILLWAY ТУРЕ
2.1. SPILLWAY TYPES AND CLASSIFICAТION (*)
It is customary to c lassify spi l lways in two main categories on the basis of the position of the inlet with respect to the full supply level in the reservoir, ie surface
spillways. the most widespread type, whose salient feature is that surplus inflows is
drawn off with only а very slight rise in water level , and submerged or orifice spillways
set wel l below ful l supply leve\, sometimes further subdivided into orifice spillways at around mid-depth, and bottom outlets.
Sш/асе spillways сап Ье further subdivided into (i) gated and (ii) uncontro l led ог f ree-overtlow typi;s. As з generзl ru\e, the open-channel tlow is steadily accele= rated from the sill ( spi l lway crest) ог weir near the upstream end, although with some tunnel types, flow may Ье free-surface at low and moderate discharges, when the control section is the si l l , and change to pressure flow over part or a l l the length of the structure at higher flows up to the ful l capacity, the control section then being an orifice or the tunnel itself. Usually pressure flow is permitted only i n the shaft portion of а conduit or tunnel and air is supplied to the free flow portion farther downstream. Air pockets must not Ье trapped in the pressure conduit as it fi l l s because this would result in undesiraЬ\e pressure variations and а reduction in the effective conduit area.
Submerged spillways normal ly have pressure flow over al l or а significant part of their length. Discharge is a lmost always control led Ьу submerged gates at the downstream end of the pressure conduit . They are, however, sometimes used for high discharges with low heads or even open-channel flow to sluice out silt after the reservoir has been drawn down to а low level or emptied ; but in this case, they should Ье considered as large bottom outlets rather than spil lways since their ful l theoretical capacity at full supply level is rarely used.
In some cases, especial ly on conduits through concrete dams, the control gate is at ог just behind the inlet, flow beyond it being free-surface flow.
There are also surface spi l lways with а short pressure tunnel at the upstream end, terminating in а gated orifice. Siphon spil l ways are а special example of this type.
Other, less fundamental criteria provide further distinction, eg :
а) whether or not most of the spi l lway is at ground level or underground;
(*) Ref. [11. 21. 27. 32. 101. 136. 145, 148. 152. 153].
1 3
Ь) whether or not there is а long chute or channel between the upstream and downstream ends (the overflow weir is one type where there is no chute) ; and
с) whether or not the spil lway terminates in а sti l ling basin or а plunge pool .
2.2. FACTORS IN SELECТION OF SPILLWAY ТУРЕ (*)
Apart from economics and danger to human l ife, the \eading factors governing the choice of the best type of spil lway for а given project are
а) reliaЬility and accuracy of flood prediction ;
Ь) seismicity of project site and reliaЬility of operation ;
с) duration and amount of spillage each year ;
d) topography and geology ;
е) dam type, and
f) operational arrangements.
2.2.1 . Flood prediction (**)
Discharge capacity under maximum reservoir level is determined from а hydrological study in which the hydrograph of the design flood flowing into the reservoir is the most i mportant factor. Dam safety with respect to floods depends on how rel iaЫe, accurate and conservative this hydrograph is. Absolute accuracy cannot exist ; it wil l depend on the amount and reliaЬility of data collected in the past on river flow, flow patterns, and meteorological aspects of major storms. А\1 other things being equal, rivers with irregular flow give the least rel iaЫe predictions, and it is only sensiЫe not to run the risk of the dam being destroyed through а relatively small underestimate of the design flood. I n other words, the spil lway design discharge, which refers to some intended maximum reservoir level, should Ье arranged to rise steeply if the reservoir level i s actua\ly higher than expected because of an u nderestimated design flood, in order to prevent, for as long as possiЫe, the dam being overtopped with all the disastrous consequences that may result.
The surface spillway, whose discharge is controlled Ьу the sill at all levels above ful l supply \evel, is better in this respect than submerged spillways because its capacity increases Ьу the power 3/2 of the head on the si l l instead of with the 112 power of the head above the centerline of the submerged spi ll way. Thus the discharge capacity of the submerged spil lway is initia\ly greater than that of the surface spillway, but the rate of increase in discharge with head is sma\ler as discussed above.
(*) Ref. [64, 65, 66, 69]. (**) Ref. [4, 10, 38, 100, 1 26, 1 39, 1 40, 1 49].
1 5
Safety against overtopping сап Ье i mproved Ьу providing an auxil iary or emergency spillway in addition to the main surface spi l l way ; it may Ье of а different type and is only expected to Ье required for the very largest conceivaЬ\e floods. I t sometimes consists of а breaching dyke lower than the non-overflow crest o f the main dam, so that it will Ье overtopped first and lower the reservoir level quickly as it i s eroded away. But caution is required because i t may suddenly release а disastrous flood into tl1e val\ey downstream, \ iаЬ\е to cause dangerous retrogressive erosion if, as is often the case, it is sited on а saddle where the geology is poor. The saddle should not Ье erodaЬ\e below the full supply level.
2.2.2. Seismicity of project area and reliabllity of operation
The seismicity of the area and confidence in the operating system are the prime factors in deciding whether it is appropriate to design а gated structure.
The impact of the seismic factor on the choice between а gated or ungated spi l lway and, in the latter case, the arrangement of the gates, is considered in section 3 .4 hereunder. Regarding rel i aЬil ity of operation, the designer must weigh the risk of one or more gates fai l i ng to open when the flood arrives because of power failure to the hoisting mechanism or а gate or gates jamming through faulty maintenance. Не must also consider the possiЬi l ity of human error in opening the gates at the wrong time, or too late because the operating rules are misconstrued. The operator must have ready access to the gate controls at а\\ t imes. It must Ье realized that an exceptional l y large flood may cause panic. lf there is the sl ightest doubt on the rel iaЬi l ity of gate operation or the competence of the operating staff, the wise choice is for an ungated spi l lway ( see I C O L D Bulletin 49).
2.2.3. Duration and amount of discharge
The cavitation and abrasion damage discussed in Chapter 5 is governed Ьу the total cumulated time the spil lway has discharged and the magnitude of the flow on each occasion. Other things being equal, the damage acce\erates with the total discharge time. All the above categories of spi l lway are subject to cavitation and/or abrasion under certain conditions so that the choice is not strictl y dictated Ьу its expected severity, but if the l ikel ihood of damage is high, precautionary measures аге necessary to retard it and еnаЬ\е repairs to Ье carried out as discussed in Chapter 5.
The duration and amount of discharge are also important factors in energy dissipation and thereby in the arrangements adopted at the point of i mpact of the discharge. The two types most commonly used, ie ( i ) the hydraul ic jump (or roller bucket) in <1 concrete stilling basin, and ( i i ) а bucket, ski jump or free overfall with а scour hole or plunge pool, pre-excavated or not, with or without protection, at the point of impact, are discussed in section 3 .6. hereunder with detai ls of the proЬ\ems inherent in both types.
1 7
The importaпce of the expected duratioп апd volume of discharge to Ье haпdled iп weighiпg the risks of damage must Ье stressed. l f the hydrological study fiпds that heavy spilliпg will occur опlу rarely for а short duratioп, the desigп may Ье Iess coпservative with l ess structural reserve streпgth i п the hydraulic basiп or l itt le ог по protectioп for the scour hole, оп the assumptioп that repairs or protective works will Ье uпdertakeп after each flood because the spil l iпg time will Ье too short for the damage to become serious.
Coпversely where the spil lway is to discharge high flows over loпg periods rhe desigпer must take serious precautioпs at the poiпt of outfall , siпce experieпce has clearly showп that very severe damage usually occurs there.
Wheп drawiпg up spillway operatiпg rules the desigпer also has to coпsider likely cases of maloperatioп that could place the dam or the dowпstream at risk. Particularly with gated spillways there is а real daпger of dischargiпg greater thaп пatural flood flows iпto the пatural сhаппеl dowпstream.
Siphoп spillways with large flow iпcreases for smal l l ake level rises may also create higher flood peaks thaп would have occured with the river i п its пatural form, for example where а l ake exteпds over most of the leпgth of а catchmeпt, thus magпifyiпg the iпflow hydrograph.
Bul letiп No. 49 is pertiпeпt to this aspect of spi l lways. Bul letiп No. 29 " Risks to third parties " сап also Ье coпsulted.
2.2.4. Site topography and geology
The topography апd geology at the site is also ап importaпt factor iп spi l lway type, ofteп iпextricaЫy tied up with опе or more of the coпsideratioпs already discussed. Опе site may Ье пatural ly suitaЫe for ап uпcoпtrolled surface spi l lway if there is room for а very loпg sill апd а relatively short chute without much excavatioп, whereas aпother may Ье more аmепаЫе to а tuппel гuппiпg partial ly or completely fu l l .
Cost, iп terms of the feasibi l ity o f usiпg spoil from the spil lway excavatioпs оп the Ьапk as dam fi l l may also Ье а determiпiпg factor iп the choice.
Chapters 3 апd 4 coпsider the impact of such coпditioпs оп differeпt types of spillways.
2.2.5. Dam type
All ог part of а spil lway сап Ье built iпto а coпcrete dam, ofteп with а substaпtial saviпg оп cost апd the possibi l ity of dischargiпg the flow straight back iпto the river сhаппеl .
1 9
Emb:шkment dams require separate spi l lway structures. The need to redirect the flow back into the river may then pose arduous proЫems .
Attempts to site the spi l lway on the dam fi ll have met with varying degrees of' success in the case of temporary cofferdam structures or at low-head dams where there is only а small head on the si l l . More research should Ье done on improving the reliability of such designs because of' their advantages of cheapness and the fact that the flow is kept in the middle of' the river channe l .
Overtopping of an embankment dam through spil lway failure or insufficient discharge capacity has immediate and dramatic consequences. This explains the greater stringency of design criteria concerning total discharge capacity, gate type and numbers, and freeboard above the assumed maximum reservoir level for embankment dams than f'or concrete dams.
2.2.6. Operational arrangements
The prospective owner of а dam project may not have experience in spil lway operation and the designer wil l have to consider client education vis а vis spi l lway type and proЫems in safe operation as well as recommendiпg suitaЫe administration, access, commuпicatioпs, etc.
\ п other cases the dam may Ье uпattended directing the decision towards an uпgated spi l lway. The \CO L\) Bul letiп 49 provides assistance under this heading.
2 1
3. SURFACE SPILLWAYS
3. 1 . COMPONENTS
Surface spil lways usua\ly consist of three parts, the sill at the upstream end which controls the discharge rate and accelerates the flow, а steeply s\oping chute where the flow velocity is maintained or increased, and the terminal structure where the flow returns to the river channel ; this may consist of а flip bucket or ski jump with а plunge роо\ or а l ined sti l l ing basin with а hydraulic jump.
The chute and terminal structure are sometimes omitted, as in some spil lways at the crests of arch dams.
3.2. GATED, UNGATED AND COMBINAТION SURFACE SPILLWAYS, Ref. [145]
3.2. 1 . Ungated spillways
Gates are optional on surface spi\ Jways. This is а distinct advantage in that an ungated spi l lway is preferaЫe when \оса] conditions such as high seismic activity, lack of confidence in maintenance and/ or operating ski\ Js , short peaking time of the inflow hydrograph, remoteness of the site and difficulty of access mean that there are doubts as to the dependabi l i ty of the gates and the way they wi l l Ье operated.
The ungated overflow s i l l is of course set at fu\l supply leveJ in the reservoir. The head required for discharging excess inflow thus means а higher dam, but this has the advantage of offering spare capacity for detaining the flood inflow above ful l supp\y level, with а consequent reduction in the maximum flow to Ье d ischarged. But special considerations such as the high cost of occasional ly flooding the reservoir rim and of the extra dam height may mean that an uncontro]Jed spiJJway is prohibitive. In this respect, the frontal s i l l whose length is usuall y l imited Ьу the configuration of the dam or topography is the most unfavouraЫe. But before opting for а gated structure, the designer must look i nto the technical and economic feasibil i ty of а Jonger uncontrol led si\J, eg, side-channel spil lway on the bank upstream from the dam, circular (morning glory) or semi-circular si\Js discharging into shafts and tunneJs, siphons, and other si \ J shapes ( duckЬi l l , daisy-shape ) . А labyrinth type of spi\ Jway crest is also useful in gaining crest Jength [61].
It has been suggested that another factor in the choice between gated and ungated spil lways should Ье the Q/S ratio, in which Q i s the peak flow оп the inflow hydrograph, in cuЬic metres/second, multiplied Ьу 3600, and S is the area of the
23
reservoir at f'ul l supply level, iп square metres . This ratio is thus а measure of' the rate of' rise of' reservoir level in metres per hour. The uпgated spi l lway is said to Ье pref'eraЬ\e iп particular if' the rate of' rise is more thaп 1 to 2 metres per hour.
3.2.2. Gated spillways
With large river Пoods апd а smal l risk of' the gates beiпg uпserviceaЬ\e or improperly operated, the gated overspi l l is general ly less costly. The s i l l is set below ful l supply level so that а high proportioп of the desigп discharge is availaЬ\e immediately wheп the gates are орепеd, if' пecessary. This епаЬ\еs the reservoir to Ье drawп dowп to the s i l l of' the gates prior to the arrival of а Пооd, this extra capacity beiпg used f'or regulatiпg outtlow апd detai пiпg all or а portioп of the tlood wave. This type of' operatioп however is поt always ассерtаЬ\е.
А surcharge capacity above f'u l l supply level is usual ly provided to iпcrease the discharge capaci ty per l iпear metre of' crest leпgth and f'urther detaiп Пoods.
lп receпt decades, spi l lways have Ьееп desigпed with coпsideraЬ\y greater discharge capaci ties per l iпear metre of s i l l with the gate fu l l y орепеd. The record-holders i пclude Karuп dam ( Fig. 1 ) iп l raп with 3 3 5 m3/s (three radial gates,each 1 5 т wide Ьу 2 1 .28 т high, with а 30 т desigп head оп the s i l l countiпg the 10 т surcharge above f'ul l supply level ).
Regardless of' how rel iaЬ\e gate operation is, it is of'teп sti pulated, sometimes Ьу пational regulatioпs, to desigп the spi l lway to preveпt overtopping of' the dam with опе or more gates f"a i l iпg to ореп. Gates сап also Ье desigпed f'or overtoppiпg. This leads to а larger пumber of' gates or the use of ап emergency spi l lway (uncontrol led overПow, breachiпg dyke or explodiпg plug) .
3.2.3. ConЬination spil lways
А combiпatioп spi l lway has ап ungated s i l l at а relatively high level апd gates set at а lower leve l . This arraпgemeпt aims at сопЬiпiпg reliaЬility with economy. Si l ls fitted with overПow gates which are опlу орепеd f'or the very largest Пoods is another combinatioп that is rare ly used. А comЬiпatioп spi l l way with ап uпgated sill at ful l supply l evel has the advaпtage of' al lowiпg relatively smal l tlows to pass without haviпg to open the gates. This сап Ье particularly usef'ul where the miпimum flow which сап Ье passed through а gate causes uпdesiraЬ\e Пoodiпg dowпstream .
3.3. INLET GEOMETRY
l пlet geometry is choseп to maximize the capacity of' the coпtrol sectioп, iп other words, the si l l discharge coef'f'icieпt. Опе also seeks а uпif'orm distributioп of' the
25
discharge along the s i l l , especia l ly if there are splitter walls ог piers, except with combination spi l lways with s i l l s at various levels .
The Creager or ogee shape i s generally considered satisfactory and the discharge coefficients obtained are well defined. Sil ls designed for subatmospheric pressure below the nappe offer only very marginal discharge improvements but with an i ncreased risk of cavitation. The best shapes for splitters and train ing walls for frontal s i l l s to obtain an even d istribution of discharge along the s i l l were found empirically а long time ago, and scale model tests аге not really necessary for every project, although they аге st i l l advisaЫe for l arge spi l lways, even those of conventional design. Scale models аге a lways needed if :
а) the valley is relatively narrow and non-symmetrica\ upstream of the i nlet with an uneven pattern of high approach velocities ;
Ь) site conditions аге such that the s i l l and pier arrangements are unusual; с) there are structures at either end of the s i l l , l iaЫe to disturb the s mooth flow
of water.
Asymmetric flow is often very noticeaЫe with side-channel and circular spil lways. The remedies required (deflectors, anti-vortex p iers) and what excavation is necessary to smooth out the current before it reaches the control section can Ье determined Ьу making adjustments in the model [78]. M odels have shown that the horizontal radius of curvature of the training wal ls must Ье not less than twice the water depth just in front of the s i l l when thi s depth is small , as is often the case with side-channel spil lways.
Long duckbi l l ог labyrinth sills require model tests to examine flow conditions and estaЫish the discharge/height rel ationship for the dam [61\. M ost side-channel spillways, being roughly paral lel to the river channel with the s i l l at the top end of the spi llway, аге tested and calibrated in the hydraulic l aboratory because the approach current is usual ly asymmetrical and they тау Ье partia l ly submerged at maximum discharge.
Model tests are also made of overspil l sections on thin arch dams to investigate vibration of the nappe and develop means of el iminating it (with spl itters for example).
3.4. GATE TYPES, ARRANG EMENT AND SELECТION, Ref. [55]
Three types of gate аге normally used
hinged fl aps or shutters ;
vertical- l ift gates; radial gates.
Flaps or shutters are only suitaЫe for modest heads of а few metres. They аге generally confined to very long s i l l s , such as river barrages. The trials with rein-
2 7
forced-coпcrete shutters that were таdе 35 years ago were пever repeated. Various types of water-operated autoтatic gates with couпterweights have Ьееп developed.
Vertical-lift gates сап Ье of very large size. They are пearly always fixed-wheel types, the Stoпey or free-roller gate поw beiпg very rare, except for very large g<:ltes operatiпg uпder high heads as used iп subтerged spil lways. There тау Ье опе or тоrе gate leaves, with or without flaps at the top. They have the disadvaпtages of requiriпg substaпtial s lots, heavy l i ftiпg plaпt апd costly, uпappeal iпg superstructures.
Radial gates are the тost f'requeпtly used for coпtrol l iпg discharge t'roт large surface spi l lways because ot· their siтple coпstructioп, the тodest force required to operate theт апd the аЬsепсе o f' slots at the side. They тау Ье as l arge as 1 5 -20 т high Ьу 1 5 -20 т wide, or 8- 1 2 т high Ьу 30-40 т wide.
А treпd that has Ьееп growiпg over receпt decades to obtaiп а substaпtial iпcrease iп discharge for а giveп gate area coпsists of settiпg the top of the gate well below Гull supply l evel апd surтouпtiпg it Ьу а тassive reiпf"orced-coпcrete water-retaiпiпg wall . This type of gate requires а top seal. Depeпdiпg оп the depth of the iпlet coтpared to the total head, this arraпgemeпt covers the whole gaтut froт the surface spi l lways discussed iп chapter 3 to the high-head апd bottoт outlets iп chapter 4.
The coпceпtratioп of the flow over а sтall width сап lead to saviпgs оп structural costs апd it is particularly useful iп пarrow valleys or where fouпdatioп coпditioпs are uпfavouraЫe. Aпother ofteп decisive advaпtage for а high-head spillway is the greater flexibil ity it offers for reservoir тапаgетепt. The reservoir сап Ье quickly drawп dowп preparatory to а flood because of the high discharge capacity at levels below full supply level .
Radial gates are usual ly preferred Гог coпtrol l iпg orifice-type spi l lways except for sтal l опеs of' less thaп 1 00 т ' ! s capacity. The аЬsепсе of side slots апd the miпiтal force required t'or operatiпg the gate give ап uпdisputed advaпtage over other types. However, vertical - l i ft gates ( fixed-wheel gates for low to тoderate heads) are поt uпusual for subтerged outlets at suitaЫe coпcrete daтs. Exaтples are Morrow Poiпt dат, USA ( Fig. 3), Ku Кuап arch dат iп Taiwaп, with four 9 х 6.6 т fi xed-wheel gates оп the cyliпdrical dowпstreaт face ( Fig. 2 ) апd Ouyaпghai dат (Chi пa) with five 7 х 1 1 . 5 т fixed-whee\ gates.
Radial gate size for surface апd ori fice appl icatioпs i s curreпtly l iтited Ьу the streпgth of the truппioп beariпgs. The latest gates are designed for тоrе than 20 MN ( 1 )
( 1 ) l MN = l Meganewton = 1 0'' N .
29
per bearing. Itaipu dam in Braz i l has 1 4 radial gates, 20 m wide and 2 1 .34 m high, applying а load of 22.7 M N to each trunnion ( Fig. 4). With Iarge gates under very high heads, the proЬ\em is generall y overcome Ьу having а series of bearings on а massive reinforced-concrete or prestressed-concrete beam, set into the training wall s downstream of the opening.
The forces applied to these beams are increasingly being transferred to the concrete on the upstream side of the gate Ьу m eans of pretensioned bars or steel саЬ\е. This greatly reduces the amount of normal reinforcement bar or structural steelwork required in the training walls .
There now seems to Ье much \ess risk of spi ll way gates jamming in seismic areas. То guard against differential displacements in the load-bearing structure, the gates may Ье set inside very stiff, one-piece frames. This makes it impossiЫe to provide contraction joints in the s i l l under the gate and in the beams, water-retaining walls or bridges above them, so that the structural Ы ocks must not measure more than 1 5-20 m from front to back or crossways to avoi d proЫems with concrete shrinkage and temperature movement. The net resul t is that there must Ье а larger number of moderately sized gates rather than а few very large ones. This also reduces risks in the event one of the gates jams in the closed position. Standby diesel-electric generators should Ье provided if power failures are l ikely.
Radial gates for surface spi l lways are usually operated Ьу wire-rope or chain winches usually attached to the leaf or to the arms behind the leaf. Ropes and chains attached to the water face of the \eaf are being used J ess often because of the danger of corrosion and danger of tloating debris becoming jammed, but are ассерtаЫе practice provided stainless stee\ or other corrosion-resistant materials are used.
H ydraulic cylinder hoists are generally preferred for orifice-type radial gates in that they сап apply downpressure on closure and help to control vibrations at partial openings. Radial gates operated Ьу two servomotors with synchronized movements have also been built for submerged and surface applications ( ltaipu, Cedillo, Agua Vermelha).
The top edges of surface spil lway gates are usually only а few decimetres above ful l supply level although there does not seem to Ье any set rule . Some discharge over the top of а gate not specifi cally designed for thi s condition is toleraЫe but i f this i s t o Ь е а normal operating condition, the gate m ust Ье built so that i t w i l l not vibrate but wi l l resist the i m pact of tloating debris and tluctuating dynamic pressures. Vertical- l ift gates faced on both sides, top tlaps and tlap gates are best for this duty.
There i s usually l ittle danger of timber jamming in spi l lways because of their large size [57). Smaller structures and orifice spil lways i n heavi ly wooded country may need а trash boom in front of the intakes although there is no rea\ assurance of successful l protection. It is better to provide wider outlets if river tloods are l iaЫe to bring down large amounts of debris.
3 1
CD··�Jl\r . .
' . " ..
.
. . . .
.. " ... r1 . .
.
г ������ � -�·
·-·
I CO m - �
®
k-
Fig. 1
Spillи:ay of Karun dam (lran).
(А) Тор of intermediate wall.
(В) Тор of end и·а/1.
(С) Tainter gates 15 х 21.28 т2. (D! Curтed buckets.
(Е) Maximum tailwater.
(F} Normal water lei·el elei·.
32
Fig. 2
Ки-Киап dam spillway (Taiwan).
33
(1) 4 orifices 6.6 х 9.3 m' each controlled Ьу fixed-wheel f/at gates.
(2) Over crest free jet.
(3) Free jets through orifices.
( 4) Pressure jets under reservoir at Е/. 952.
(5) Original apron.
(6) Additional apron.
0
в � в -1 �-
д --1
.1
SECTION В-В SECTION А-А
1 0 " '
Пg. 3
Elei·arions ШI{/ sections of" rhe spi//1i·a_1- orifices оп 1/1е Mormн· Poinr Jam (l!SЛ).
0
( 1) Upsrream elei·arion о( spil/11·a_1· х 4.58 т 2. i(!fal capaci1_1· 1 130 m 3/s_
(2) Section tl1rouкh outsi1Je orifice.
(3) Secrion 1hro11кl1 inside ori(ice_
34
orifices 4.58
Л10fi' H' Poit dam (SA).
35
0
0
0
�25.00_
V U E E N P L AN PLAN V/E W
®
14. 945 %
i = 1 7. 6 33 %
i = 17. 6 33 %
P R O F I L PROFIL E
.. ----·
Fig. 4 Itaipu dam (Brazil) : Spi/lway and spil/way gates (see photograph, page 46).
4 - 1
3 6
А . Р/ап view.
(1) Left chute.
(2) Dividing wall.
(3) Central chute.
(4) Right chute.
В. Longitudinal profile.
(5) Road tunnel.
(6) Air slots.
)3700
183.75
4 - 2
3 7
С. Spil/way gates.
(1) Weir axis.
(2) Control room.
(3) Access platform to servomotor articulation.
(4) Trunnion girder.
(5) Trunnion thrust Ыосk.
(6) Tainter gate.
(7) Servomotor.
(8) СаЬ/е and hydraulic pipe gallery.
(9) Stop/ogs.
( 10) Gantry crane for lifting stoplogs.
3.5. CHUTES
The chute downstream from the sill conveys the water to the point of outfall downstream of the dam. Its length and shape will Ье dictated m ainly Ьу dam type and site topography.
Concrete dams have the advantage of offering а ready-made load-bearing structure for the sill and chute, although there тау Ье obstacles, eg, i nsufficient dam width for the required discharge capacity, congestion, or serious risks of not being аЬ!е to dissipate the energy c\ose to the dam's toe. But comЬining the spillway with the dam is generally the most economical arrangement, and it usually keeps the jet aligned with the river channel so as to minimize erosion of the bank downstream of the stil l ing роо\.
Chutes for overtlow spillways on concrete dams usually convey free-surface flow, a\though there is sometimes а short pressure tunnel at the upstream end where the orifice-type gates аге housed. Where the spillway is adjacent to or separate from the dam, the chute тау have а variety of configurat ions, wholly or partly in the open and/or underground. The determining factors in the choice are the geology, the shape of the valley slopes and the configuration of the downstream watercourse.
Thin arch dams with overtlow sections at mid-crest usually have practically no chute. They тау have а free-falling nappe or а jet under а very low head and discharging only а very short distance from the toe of the dam. This arrangement is suitaЫe for small and moderate d ischarge capacities ; it requires а certain width of val\ey at the dam toe and some minimum depth of water in the impact zone compatiЬ\e with the height of fall and discharge per metre width. Arrangements such as splitters are frequently made for breaking up and aerating the nappe to reduce its erosion potential in the i mpact zone. То Ье effective, such devices must Ье set where the tlow velocity is greater than 1 О т! s and must Ье about as thick as the nappe, and this sometimes means that а short concrete chute i s required for а height of 5- 1 0 т [ 123].
Where the overtlow section takes up а large part of the crest of а thin arch dam, it is often necessary to arrange splitters on the sill in order to prevent vibration of the free falling nappe being set up generally for relatively low discharges which cou]d adversely affect dam. Such precautions have been m ade on Zaouia N 'Ourbaz dam in Morocco and Hautefage and M oulin Ribou dams in France [79\.
Surface spillways on mass concrete dams and on or i n one of the abutments require а channel of significant length, wherein the tlow accelerates.
39
А surface spi l lway on the abutment adjacent to the dam is an interesting proposition because of the moderate length of the channel required. H owever, for high discharges, it often encounters difficulties with the amount o f excavation required, danger to bank stabi l i ty, and the angle the flow makes with the river channel.
With the side-channel spil lway, some of the energy is dissipated in the gently sloping top end of the chute under the sill. It is com mon practise to control the level in this first stilling basin Ьу means of а second, much narrower sill or side contractions whose height is careful ly calculated, marking the start of the steeper part of the chute (spi l l way at Youssef Ben Tachfine dam, M o rocco, Fig. 5) .
U nder favouraЫe conditions where the val ley m akes а sharp bend or there is а tributary val ley near the reservoir, this feature can Ье made use of to keep the chute short. It must however Ье recognized that discharging into а secondary val l ey often causes trouЫe because of the greater erosion potential at the outfall than in the main valley, and environmental proЫems m ay Ье caused Ьу discharging large volumes of water into а previously near-dry valley.
As а general rule, chutes and outlet tunnels are straight because they are designed for super-critical free-surface flow that is difficult to turn. I deally, subcritical flow is enforced where there are changes in direction with the control or critical section at the top of а straight chute.
Large-radius bends are ассерtаЫе in super-critical flow provided that the resulting shockwaves are contained within the chute. It is easier to turn pressure flow, which is why some recent designs have а curved pressure tunnel at the upstream end of the spillway where the head is small, with а control orifice at its downstream end where the straight chute starts.
Where turning of super-critical flow must Ье adopted, model testing for the ful l range of flows i s necessary. N o t only i s the shock wave pattern l ikely to vary, but for all flows other than the one for which the bend is designed, there may Ье asymmetry and/or sinuousity downstream causing а widely varying performance of а bucket or flipper and possiЬly of а still ing basin.
If the chute is relatively long, а gradual reduction in width commensurate with the acceleration of the flow can produce some savings on cost. But like piers and horizontal curves, it m ay produce standing waves. At l arge dams it is wise to run scale model tests to determine the heights of these waves and their influence on the transverse distribution of flow at the point of junction between the chute and the terminal structure and to optimize the configuration.
The chute may alternatively widen gradually to reduce the flow per unit width downstream, reduce the depth of water and cavitation damage risk, and improve energy dissipation.
4 1
G о Е о о � о (\J -
о (Х)
о <.О
8 о v t-о z (\J N
1-о 1
а:: Q <(
е (!) -
<( -
� ---
42
Fig. 5
Youssef Веп Tachfine dam (Morocco) : Surface spillway.
43
(!) Longitudinal profi/e.
(2) Road bridge.
(3) First contro//ing si//.
(4) Second contro//ing si//.
(5) Bottom outlet and outlet of diversion tunnel.
(6) Dejlector bucket.
(7) Р/ап vie»·.
0
�Е -/
®
L 30 50 j_ 70. 10 l
44
4
17_5. 2 5
�Om
о 10 20 m --.....
Fig. 6
La Grande 2 dam (Canada) : Spillway.
(А) Plan view
(В) Longitudinal proflle.
(/) Approach channel.
(2) Spil/way.
(3) Upstream stilling basin.
(4) Steps of the unlined chute.
(5) Main dam.
(6) Submerged dike.
(7) Natural ground profile.
(С) Proflle of the upstream stilling basin.
45
( 1) Gate operating tower.
(2) Stop/og s/ots.
(3) Service bridge.
(4) Fixed wheel gate.
(5) Guiding walls.
ltaipu dam (Hrazi/).
46
's).
La Grande 2 ;/ат (Canada). Sp1'11�ay ( 16 140 m1/s).
47
Discharge tunnels are suitaЫe for narrow val\eys rising well above the dат crest because the construction area is kept separate froт the тain works, а substantial advantage at congested sites. The Jongitudinal profiles of outlet tunnels are affected Ьу whether or not they were originally tunnels for river diversion during construction, Ьу the inlet shape (fronta\ or side-inlet si \ I , circular or seтi-circular si l l , etc.), project purpose and geo\ogy. The upstreaт end is usually vertical or steeply sloping to acce\erate the flo w ; every effort is таdе to keep this drop to not тоrе than 60 тetres to control the cavitation risk (Chapter 5) along the reтainder of the tunnel, where the gradient i s just steep enough to keep the water flowing [52\.
In order to тiniтize excavation, it is soтetiтes accepted that beyond а certain discharge, the upstreaт end of the tunnel тау run ful l , the control section then Ьесотеs the orifice at the end of the pressure section instead of the inlet sil l . This arrangeтent involves the рrоЫет of entrapped air pockets already тentioned and also has the disadvantage that little extra eтergency capacity is availaЬ\e, since discharge will not great\y increase as the headwater Jeve\ rises. There тust Ье abso\ute confidence in the design flood hydrograph therefore, especially if the dат is an eтbankтent structure . l t is only al\owaЬ\e for the pressure section to extend downstreaт of the grout and drainage curtains under the dат if the rock is exceptionally watertight and coтpetent and i f there is а coтfortaЫe rock cover above and to the sides of the tunnel. But it shou\d generally Ье avoided.
Outlet tunnels running partially ful l always raise а рrоЫет of aeration at al\ discharges. Turbulence is the cause of air entrainтent in the forт of bubЫes scattered throughout the flow. l t is соттоn practice to Jeave 20-35 % of the tunnel area етрtу above the waterline as sufficient тargin against pressure flow, the discharge area being calculated without allowance for air entrainтent. lf freesurface flow extends the who\e Jength of а short tunnel there wil l Ье а natura\ draught above the water surface. But if the tunnel runs full at the upstreaт end, the air entrained Ьу the water сап only Ье rep\aced froт the downstreaт end, flowing in the opposite direction. With l ong tunnels and high flow ve\ocities, the usua\ 20-35 % free section тау Ье inadequate and а vent will then Ье provided just downstreaт froт the section runing ful l . Measureтents on тапу daтs, Jaboratory research and theoretica\ analysis have produced equations re\ating air flow to water flow for various flow conditions, providing an orientation for design criteria for selecting adequate vent size [50, 82, 83, 137\.
Hydrau\ic juтps in the near-horizontal part of tunnel outlets are avoided as а general rule, the governing factor being the height of this part with respect to water levels in the natural streaт at the outfa\J . H owever, а hydrau\ic juтp in а tunnel followed Ьу Jow-pressure flow was recently proposed as а тeans of energy dissipation for the bottoт outlets at the Souapiti scheтe on the Konkoure river in Guinea. The position of the change of flow regiтe is controlled Ьу appropriate
49
geometry, and ample protection is provided against tluctuating upl ift pressures and cavitation.
I n somes cases, particularly favoraЫe geology (hard, unjointed rock) has made it possiЫe to \eave part of the chute unlined (excavated as а series of steps at La Grande 2 ( Fig. 6), James Вау) [ 1 2] ; protection is sti l l required over fissured zones. Such designs require very careful study, as retrogressive erosion of the channel can start at the discontinuities and progress very rapidly ( Ricobayo dam in Spain [42] ) .
At high chute velocities (30 m/s and more), the most serious proЫem is cavitation damage, especially with the normal type of free-surface tlow. Cavitation and modern methods of preventing and combating it are dealt with in Chapter 5.
3.6. TERMINAL STRUCTURES AND ENERGY DISSIPAТION
3.6. 1 . PreamЬ\e
The crucial issue at the point where spi l lway discharge returns to the river channel is energy dissipation. Under natural conditions without the dam the energy of the water was dissi pated linearly Ьу friction and turbulence, the result ing erosion being spread out over the whole reach of the river and its tributaries subsequently contained within the reservoir. Once the dam is built , most of this energy must Ье dissipated at а single point where al\ the erosion potential i s concentrated, ie, the point where the spil lway returns the water to the river.
Although the l i terature reports an increasing number of large spil lways claimed to Ье great successes in terms of design and construction, the actual performance record is sti ll short since most have only operated at much smal ler tlowrates than their design capacity, and for very short periods of time. The smaller number of spillways that сап boast prolonged operation at high tlows like Tarbe\a [87] ( Fig. 7) have been the sources of frequent, serious trouЫe. This implies that, except where the energy of the water is converted to e lectricity, the proЫem of energy dissipation is still not totally under contro l . The larger number of increasingly high dams being built on large rivers with irregular tlows m ust Ье а serious concern for present and future generations.
The energy is almost always dissipated Ьу а sudden slowing down of the tlow and а concomitant change from super-critical to sub-critical conditions. The engineer can achieve this in one of two ways :
а) The whole distance over which tlow is super-critica\ may Ье canalized, and the change conf'ined in an art ificia l structure, the hydraulic jump basin whose geometry is determined from theory and experiment.
5 1
о 5 0
1 09,80
® Fig. 7
1 0 0
· - 7 9 ,3 0
d� - 'i. _T U!'�
r<1
1 5 0 m
-----+ - - - -��NN� - -�f-+--
12 1
Tarbe/a dam (Pakistan) : Stilling basins of tunnels Т 3 and Т 4.
52
(А) Р/ап view.
(В) Longitudinal profi/e.
(1) 2 sector gates 4.88 х 7.32 m2 under 136 т тах. head.
Tarbela dam f Pakista11). Тит1е/ 110. 4 at maximum flow after replaci11g 1l1e l1ydraulic jump basin Ьу 1"'0 cha11nels 1н1/1 flip bucket.s.
53
Ь) Alternatively, the turbulent tlow тау Ье thrown up into the air above the river channel as а jet or nappe with some degree of horizontal ve\ocity. The change in tlow conditions occurs in the scour hole or plunge pool. А scour hole is suitaЬ\e if it forms natural ly with the fi rst spil ling and therefore grows s lowly enough so as not to endanger the stability of vital structures or the nearby valley s\opes over а reasonaЬ\y \ong period of time.
But it is becoming more frequent to excavate al \ or part of the hole so that it can Ье concrete-lined in places where erosion is not ассерtаЬ\е . It might also form the quarry for excavated rock used for part of the dam's construction.
For small design discharges, the nappe may Ье spread out sideways to make it thin enough for the spl itters and aeration devices described in section 3.5 to Ье fully effective. Most of the energy will Ье dissipated Ьу friction between the air and water, ant the plunge роо\ wil l Ье of only marginal importance (see " Free-fal l ing nappe" in section 3 .6.3 .) .
3.6.2. Stilling basins : accidents, hydrodynamic proЫems, uplift pressures, vibration, cavitation, abrasion, maintenance
Hydraulic jump
At first sight, dissipating the energy in а hydraulic jump appears very attractive. It is confined within а l imited volume whose length and location are theoretically well defined, and the remaining energy leaving the basin can Ье calculated quite accurately.
It is for these reasons that the hydraulic jump has always been and remains the hydraulic engineer's favorite, because the wishes to design, Ьу theory and experiment, the best shapes and dimensions to contain а short, staЬ\e and efficient standing wave. The abundance of literature on this subject means that the designer has no proЬ\em in turning up ru\es and precedents to suit his own project. The question of dimensions is one of pure hydraulics and wil l not Ье considered here. I t must however Ье remarked that the jump is capricious and confining it to the desired position inside the basin or at the end of the chute (as with the rol\er buckets on some overtlow gravity dams) depends not only on the relative tailwater \evel but also on minor changes to the geometry of the tloor and training wal ls . А sharp discontinuity in channel shape (eg, а vertical step) which generates а large force in the upstream direction helps to l imit the possiЫe movement of the jump under differing conditions. Substantial safety margins must Ье taken to prevent the jump from escaping, especially in relation to the variation of the tailwater level as intluenced Ьу degradation or aggravation downstream of the basin. Degradation of the topography downstream of the dam could lead to eventual tailwater recession and а once satisfactory stilling basin could cease to function properly. Some end sill fixture would preclude this. Another alternative is to use the submerged rol ler bucket placed at а sufficent low level to get one roller forming in the bucket whi\e another one is turning in the opposite direction immediately downstream of the dam toe. This alternative has been used successfully many times in USA and I ndia, apparently
55
without any damage but for discharges and times which are not known (cf. I ndian National Committee on Large Dams : " Water Resources Research in I ndia ", Chapter IX, New-Delhi, 1 979).
The most serious proЫem with the hydraulic jump dissipator however is more one of structural strength than hydraulic efficiency. The following section discusses some relevant ideas.
Firstly, there may Ье no practical alternative to а hydraulic jump if site conditions make prior excavation or subsequent erosion unacceptaЫe, especially in the case where the spil\way is expected to operate frequently and for !ong times.
Experience from recent decades gives many examples of stil l ing basins suffering serious damage from prolonged flows approaching maximum discharge capacity. This happened with the spillway at M al passo dam in M exico for example [ 1 29, 1 30, 13 1 , 1 32] ( Fig. 8), and the i rrigation tunnels at Tarbela dam in Pakistan deserve special mention, even though they had nothing to do with surface spillways, because of the very high flow and head (maximum discharge 3 ООО m 3/s per basin, head between 90
-and 1 40 m), their very long periods of operation (several consecutive
months each year) and of course the puЬ\icity given to the fr�quent trouЬ\e that occured ( 1 974- 1 975- 1 976).
The most frequent type of damage recorded is complete jloor slabs being torn ир, which may or may not Ье followed Ьу erosion of the foundation (this went 25 т into the underlying rock at Tardela) [86, 88]. This i s а clear sign of high uplift pressures under large areas of the floor. Although this is the major factor, others may favor the appearance of uplift pressures and aggravate the damage : cavitation, abrasion and vibration.
Uplift
The uplift pressures tending to l ift the slabs are caused Ьу the intermittent conversion of kinetic energy into pressure energy through any openings there m ay Ье in the channel floor. This mechanism poses а threat especially at high Froude numbers and is accentuated Ьу intense turbulence, ог macroturbulence, Ьу which the energy of the water is dissipated in the hydraulic jump. А t'eature of macroturbuleпce is the high frequency pressure fluctuations of widely varying amplitude at al\ points in the flow including the wetted surfaces of the invert and training walls [22, 1 3 1 , 132, 1 5 1 , 1 64]. 1 11 t h e worst case, the half-amplitude i s close t o 0 . 4 (V2/2 g ) in which V is the mean inlet velocity of the turbulent flow. The frequeпcy is around 1 H ertz.
Statistical analysis of pressure records from surface-mouпted instrшneпts give а mеап square raпge for dynamic fluctuations of 0. 1 О to 0 . 1 2 (V212 g).
5 7
1
о о
�·
8 "' N
L9' LSE+O Н3
о о N N � �
г E8'86 l +О 153
/
о о
о � -
58
1 .
8. "' N
.. ···
{ .
Fig. 8
Malpasso dam (Mexico) : Plan view and longitudinal profile of spil/way.
5 9
( 1) Original slabs.
(2) Spillway р/ап.
(3) Erosion damages.
(4) Spil11vay /ongitudinal cross section.
(5) New anchor bars.
(6) Spillway transversal cross section showing erosion damages.
(7) Spillway longitudinal section showing erosion damages.
(8) Stilling basin plan showing reconstructed zone (shadoived).
When the pressure Ьесотеs negative at а point on the invert, one тау have а short l ocal instabi l ity if there is а steady uplift pressure at the concrete-rock contact or at any other interface within the thickness of the s\ab (horizontal construction joint) or foundation and this uplift is greater than the subтerged weight of the overlying rock or concrete plus the water pressure at that tiтe on that point. The total upl ift force тау Ьесоте dangerous if there is а steady uplift pressure under large areas and negative Пuctuation of sufficient aтplitude appear siтultaneously on sufficiently large areas of the invert. Daтage to тапу sti l l ing basins indicates that the probabil ity of occurrence of this unfavoraЬ\e coтЬination is far froт being negligiЬ\e.
The build-up of uplift pressure soтewhere under the invert, especially at the rock-concrete interface, while the basin is operating, is тоrе the rule than the exception. The pressure is generally governed Ьу the tailwater level behind the basin, and i t тау gain access through the drainage systeт, which is itself required because of the nearness of the reservoir or the water tаЬ\е in the bank overlooking the basin. If the required depth of water for the juтp is at the end of the still ing basin Ьу an end sill control , а gravity outfa\I froт the drainage systeт тау Ье found within а reasonaЬ\e distance. But the depth of the sti l l ing basin Пооr is тоrе usually governed Ьу natural Поw conditions in the downstreaт watercourse and тust Ье set well below river bed level, тeaning that gravity drainage is not possiЫe and also that tai lwater pressure may enter through the drainage systeт, if it is not isolated. An independent drainage systeт with puтps to prevent the build-up of pressure is of course the best conceivaЬ\e assurance against uplift froт negative pressure Пuctuations. But pumping is expensive and not always rel iaЬ\e ; i t is only used in the extreтe case of l arge structures operating continuously (sti l l ing basin to No. 3 tunnel at Tarbela after the third set of repairs) and at hydro power projects where station attendants are availaЫe for рuтр тai ntenance.
Without а puтped or gravity drainage systeт, the uplift force under the invert at any given point and time сап Ье near the tailwater head. The very unequal spatial distribution of negative Пuctuations at any tiтe over an area of the i nvert тeans that the total тaximum uplift force under this area is distinctly less than the uplift pressure тultiplied Ьу the area. Russian experiтents have yielded equations giving the total uplift force under а rectangular Пооr slab as а function of the head, spil lway discharge per тetre width, slab length and slab position. Each slab is assuтed to act independently. These equat ions reveal the consideraЬ\e attenuation of effective теаn uplift pressure with slab length and distance (in the downstream direction) froт the standing wave front ; if these equations were to Ье fol lowed strictly, i t would often теаn using relatively thin Пооr slabs not tied into the foundation.
E xperience however counsels prudence. Firstly, the position of the wave front is not accurately known, and тау move. Secondly, instaЬil i ty conditions other than
6 1
those caused Ьу negative pressure Пuctuations and the downstreaт upl ift pressure тау arise. This is unlikely, but they аге l iaЫe to produce higher upward forces on the slabs. It is not inconceivaЫe that а positive pressure Пuctuation could Ье transтitted тoтentari ly under а significant portion of the invert through а geological joint ог open crack between the water-filled drainage systeт and the inside of the basin ; ог that in the sате way, an upl ift pressure equal to V2/2 g (the velocity near the wetted surface) building up on а fortuitous projection into the current (caused, say, Ьу а vibration of а s\ab ог а spal led joint) could penetrate into the structure. This leaves us with an uncertainty as to the highest тagnitude of the force trying to l ift the slab. The тiniтuт precaution would sеет to Ье to design the invert to withstand the тоге severe of the two situations :
а) fu l l downstreaт uplift pressure appl ied over the whole area of the Пооr with the basin етрtу, or.
Ь) Full upl ift pressure equal to the теаn square of the тacroturbulent pressure Пuctuation (0. 1 2 V2/2 g, with У being the inlet velocity) applied under the whole basin, with the basin fu ll .
If there is enough cohesion in the foundation, plain ог prestressed anchors can contribute а substantial part of the necessary strength ; otherwise the invert тust Ье made thicker and if possiЫe held in place Ьу the side walls.
Various constructional arrangements can Ье recoттended to oppose uplift daтage to the Поог fгот these macroturbulent pressure Пuctuations :
а) All contraction joiпts should Ье fitted with properly located and eтbedded seals.
Ь) There should Ье no drain openings in the training wal l inside the basin even above the " theoretical " water line upstreaт of the standing wave. H owever, drain outlets in а dentated s i l l at the beginпing ot' а sti l l iпg basin have proved to Ье satisf actory.
с) Кеер the areas of the Пооr s labs as large as possiЫe.
d) Coпnect slabs Ьу means of dowels, shear keys and reinforceтent across thejoints.
е) Кеер horizontal construction joints to а тinimuт, with dowels across theт .
./) I f drainage is real l y necessary, keep it well away ( 1 т to 1 . 5 т at least) fгот the wetted surfaces so that abrasion ог cavitation wear wil l not make it accessiЫe to the macroturbulent Поw.
I n the same coпnection, it тust Ье reтarked that leaving the basin unliпed (as at the spil lway at Раlота dат iп Chile which is coпcrete lined Ьу now) on the argument that the rock is of exceptionally good qual ity would Ье hazardous. However good а rock тight Ье, i t always contains cracks so that the process of destruction Ьу uplift pressures тoтentari ly unbalanced because of dynaтic Пuctuations (wedge effect) would Ье particularly effective and destructive in the absence of any protective l ining.
63
Vibration
I n the macroturbulence method of energy dissipation, the dominant pulsating components (those with the greatest amplitude) have frequencies between О and 1 0 Hertz. This means that some parts of the basin like the invert slabs and any deflectors or splitters it might сапу are in danger of resonant vibration (231 ( cf. measurements on slabs in the Tarbela irrigation tunnel basins before they were strengthened and anchored). Vibrational movement of the slabs opens interfaces in depth, and throws up the corners of the joints, encouraging dynamic uplift and the lifting of the s\ab.
Vibration prevention also requires massive slabs, connected together Ьу reinforcing bars across the joints and/or shear boxes. They should always Ье pinned to the foundation where possiЫe.
Lastly, these pressure fluctuations сап produce fatigue in some structures (especially rock anchors) that is difficult to estimate.
Cavitation
Pressure тау momentarily become subatmospheric in а part of the hydrau\ic jump because of these macroturbu\ent pressure fluctuations causing cavitation. This is however opposed Ьу the very high degree of aeration of the flow, again caused Ьу the macroturbulence. Deflector Ыocks and other obstacles in some basins are very exposed to cavitation erosion. The end of the chute and the upstream lip of the basin under the leading edge of the standing wave are particularly sensitive areas because the high velocity turbulent flow is in intimate contact with the bottom (except in some configurations with а step or bucket between the chute and basin). Cavitation erosion is therefore а frequent occurrence in these still ing basins. Although less sudden and spectacular than uplifted slabs, it is nevertheless the cause of repeated damage. lt would hardly seem possiЫe to eliminate this damage completely because of the nature of the flow in the basin. Submerged Ыocks of the Rehbock or US Bureau of Reclamation (basin Туре 1 1 1 ) type are particularly at r isk for velocities exceeding 1 5- 1 8 m/s ; it is only possiЫe to slow down the damage Ьу using special concrete or steel l inings ( 1 20, 1 2 1 , 1 22].
Abrasion
The last cause of damage is abrasion. With surface spillways, the risk of abrasion from sediment carried Ьу the flow is nonexistent or at worst will not occur for а long time, until the reservoir has silted up to sill level. Bed \ oad and suspended sediment is dangerous inasmuch as i t contains а significant proportion of hard particles the size of fine - to medium - grained sand (e.g., angular quartz grains).
Some basins are damaged Ьу abrasion from bed load (alluvium, rubЬle) from the river channel just below the dam [34]. This wil l happen if the basin is too short
65
or badly shaped, so that there are back-curreпts over the uпprotected river bed at the basiп outlet. Abrasive sedimeпt апd coпstructioп debris left оп the basiп iпvert before it is first fil led or rocks throwп iпto the basin Ьу visitors сап Ье trapped iпside the basiп апd cause consideraЬ\e wear, especially where there is а sharp outlet sil l . Stilling basiп covers апd feпces may Ье necessary to preveпt rocks from beiпg thrown iпto the basin. Accideпtal asymmetric distributioп of the turbuleпt flow оп the iпlet chute is always ап aggravatiпg circumstance for abrasioп because it coпcentrates wear on certaiп areas of the wetted surfaces.
For protectioп agaiпst abrasioп from trapped sedimeпt, the basiп m ust have а self-cleaniпg shape, so that sedimeпt left in р\асе or washed in from time to time from upstream or dowпstream is quickly expelled. This сап Ье easily checked оп а scale model [85]. I п actual field applicatioп self-cleaпiпg is hard to achieve. То prevent m ovement of the dowпstream bed iпto the basin, traps at the епd of the basiп may Ье required.
If the flow a\ways carries abrasive sedimeпt, wear is iпevitaЬ\e. E xperieпce shows (Sап Меп Xia dam оп the Yel low River) that there is little wear оп flat concrete surfaces parallel to the flow еvеп with high abrasive saпd loads (50 kg/m3)provided the velocity V is less thaп 1 0 m/s. Веуопd this, the rate of wear iпcreases as V3, iп other words very quickly. Velocities iп excess of 1 О m/s are commonlyfouпd at the iпlets to stilliпg basiпs апd persist for some distaпce iпside. The turbuleпt flow within the jump also tends to aggravate abrasioп.
Special concretes апd liпiпgs developed to date slow dowп abrasion wear but never dispense with it completely. Iп апу eveпt, they would Ье far too costly for а stil l ing basiп in view of the areas involved, except of course for smaller structures.
Repairs
The three processes Ьу which stilliпg basiпs are damaged, ie, uplift, cavitation and abrasion, сап соmЬiпе, one beiпg l iaЬ\e to cause the other. The net result is that damage speeds up. This is why the hydraulic jump basiп is а very vulпeraЬ\e structure.
Because of this, the desigп must include arraпgemeпts for quick repairs once serious dam age is detected. The most commoп difficulty is опе of dewatering the basin. The floor is almost always set well below tailwater level to obtaiп the couпter-head пecessary for coпtaiпiпg the jump within а structure and unless the river runs completely dry, some sort of cofferdam is пecessary. It сап Ье iпserted quickly if it coпsists of stoplogs beariпg against piers and handled from а gaпtry. But it may well Ье very expeпsive.
If the spillway is expected to operate ofteп for proloпged periods, а good precautioп is to divide the chute and basiп up lengthwise so that work сап Ье carried out еvеп wheп it is operatiпg.
67
3.6.3. Ski jumps and nappes, scour holes, erosion, aprons, hydrodynamic proЫems
Ski jump (jlip buckets)
The terms " ski jumps " and " fl ip buckets " suggest deflectors at the end of а chute. These detlectors impart а sudden change of direction on the shooting flow from а concrete structure. The jets are thrown up into the air and hit the river channel at а predetermined spot. Energy dissipation occurs chiefly in the water around the impact area. This terminal deflector, also known as а bucket, is set above tailwater level. I ts height, lip angle and total head are the main parameters determining the trajectory of the jet. The jets themselves can also impact, aiding in the dissipation of energy (see below l ntersecting-jet spillways).
The airspace between the nappe and the tailwater has been found to Ье an important factor. Should this airspace Ье too limited, the nappe can Ье drawn down, thus setting up trouЬ!esome oscil lating flow conditions (possiЫe dam vibration).
Sediment bars which form the eroded material from the scour holes can raise tailwater levels thus reducing the airspace to the nappe. This reduction may give rise to the undesiraЫe flow osci l lation proЫems mentioned above.
The designer's main concern is usual ly to have the impact zone as far as possiЫe from the bucket to protect the fou ndation against retrogresive erosion. This is obviously а difficul t task for low heads (less than 50 m). As regards the position of the impact zone relative to the valley banks and the shaping of the jets to skew them, spread them, break them up, etc., each site is а separate proЫem sui generis that can only Ье properly examined on а scale model. This is the reason behind the great diversity of fl ip buckets : they may Ье symmetrical or asymmetrical, they may Ье shaped as facets or progressively curved surfaces, etc. One can see various shapes for the same general situation whose hydraulic performance is not signifi cantly different .
One special case concerns buckets set very low at the downstream portals of diversion tunnels subsequently converted to act as spi l l ways. In some projects (Sidi Mohamed Ben Aouda* in Algeria), these buckets are submerged at high flows and should then more strictly Ье termed roller buckets, as they produce а type of hydraul ic jump ( Fig. 9).
Free-falling парре
The free-fall ing nappe, without chute, is really found only on arch dams whose downstream face is overhanging or near vertical .
Some of the energy can Ье dissipated before the nappe reaches the river. Splitters and deflectors can Ье placed on or а few metres below the sill so that the nappe is divided into smaller jets which impact the adjacent jets or against deflectors before shooting. This increases the friction area between air and water along the
• Recently named Es Saada.
69
Fig 9 .
Spi/lway or s ·
d·ment
� 1 1 м h (recently пате
/ E
aтS
ed Веп Aouda ds aada) (А/ . ат
�-
""").
�, lo ( 1 ) General plan view pf project
88. 50
0r �
(А) Low lwvwl downstream deflectors
(2) Spillway longitudinal profile
70
50 lOOm
10_-+-б_О_m_
1 54.00
137 50
о бОm
(3) Cross section of entrance tower.
А . А
12 .50
(4) Cross section of galleries underneath the embankment.
7 1
о А д
5 Ю m
Fig. 10
1
11
1 1
1 1_j_ _ _ _ _ __ __L
в в
Hendrik Verwoerd and Р. К. Le Roux dams (South Africa) Splitters of surface spillways.
А-А Vertica/ section. В-В Downstream elevation. (1) Splitter. (2) Deflecting p/atform. (3) Aeration vent.
72
Р. К Le Roux dam (Soutl А/пса), Uncontolled spill a и:ith splittes.
73
trajectory. The reduction in the throw is а measure of the effectiveness of this friction effect. Dissipation will only Ье sigпificaпt i f fragmeпtatioп produces а marked emulsifyiпg effect ; iп practice this is only really feasiЫe with thiп парреs (less thaп 5-6 т head on si l l ) , ie, for small flows of higher frequeпcy.
The splitters desigпed Ьу D. F. Roberts in South Africa апd flip buckets built iпto the overspill s i l ls of arch dams iп that country [49, 1 23] ( Р. К. Le Roux, H eпdrik Verwoerd [ 1 3] оп the Oraпge River) meet the above requiremeпt ( Fig. 1 О). AppreciaЫe dissipatioп is obtaiпed in the iпteпsely aerated flow wheп the valley is wide eпough for the парре to Ье spread sideways (specific discharge at Р. К . Le Roux апd H enrik Verwoerd 60 m3 /s.m approx., theoretica] парре thickпess at splitter level 4 m). Professor Aubert's system of staggered bottom-hiпged gates produces а similar effect [ 1 1 ] .
Iп most cases however most of the kiпetic eпergy is stil\ availaЫe at the poiпt of impact, апd is suddeпly dissipated iп the surrouпdiпg water. There is по fuпdamental differeпce with coпditioпs iп а hydraul ic jump ; eпergy is dissipated Ьу macroturbulent frictioп апd the characteristic coпcomitaпt effects (dyпamic pressure fluctuatioпs, cavitatioп, abrasioп) are the same.
The chief differeпce is in the fact that а larger receiviпg volume of water is iпvolved апd that the greatest turbuleпce is поt iп such iпtimate coпtact with the wetted surfaces. Situatioпs where there is а small volume of water are traпsieпt or exceptioпal ; they are sometimes eпcouпtered оп the first use of the spillway where there has Ьееп по prior excavatioп of the scour hole or at the impact of the парре with а coпcrete арrоп iп shallow water.
Scour holes (5, 93, 94, 96]
The use of а flip bucket is based on the belief that the depth апd area of the scour hole will поt iпcrease sigпificantly with contiпued use of the spillway Ьеуопd а c_ertaiп degree of developmeпt.
Various authors have put forward equatioпs to give the maximum depth of the scour hole d iп metres be/ow tailwater /evel in ап isotropic graпular material, produced Ьу а free-fall ing nappe from а coпtroll si l l . The most significaпt parameters are the uпit discharge q iп m3/s.m апd the tota\ head h iп metres from the reservoir level to the water level iп the scour hole. The best kпоwп of these equations is from Veronese which yields the average va\ue fouпd from scale-model tests :
d = \ .9 hu ш q0 54
The Martins equation claims to give the eпvelope values of observatioпs from scale models апd 1 8 full-scale dams *
d = 2.3 hu 1 0 qo о
• In the Martins equation [94], h is the difference in height between the reservoir level and (i) the water level in the scour hole for а free-falling nappe from the control sill or (ii) the lip of the flip bucket for а spillway with chute. Martins states that in all cases examined this lip is only а few metres above tailwater level.
7 5
Other equatioпs are giveп Ьу Robert L. George [49]. They are a l l iпdepeпdeпt of the type of material iп which the scour hole forms because they incoгpoгate the assumption that the final depth is indepeпdeпt of this parameteг, cohesioп and hardпess опlу affectiпg the time гequiгed fог а staЫe depth to Ье reached. The value of these equatioпs fог an oгdeг-of-magnitude estimate is гecognized Ьу designeгs, еvеп iп the case of П iр huckets at the епd of loпg chutes, despite the fact that the aпgle of impact which the jet enters the pluпge pool and tгansverse distгibutioп of the jet will Ье diffeгeпt апd аге importaпt paгameteгs поt takeп into accouпt. These equatioпs give по diгect infoгmatioп on the pгofi le of the scour hole ог, of course, on апу retrogressive ог sideways eгosion (back-curгent) or the pгactical coпsequeпces thereof.
At best, опе сап hope that the final upstгeam developmeпt of the scour hole will Ье reached without any seгious dапgег to the stabi l ity of major stгuctuгes, еvеп though dowпstream eгosion of the гivег channel апd baпks гemaiпs uпcoпtгol\ed. Such а favoraЫe situation is гаге, i t гequiгes а wide valley betweeп low, geпt ly slopiпg baпks. The valley is usual ly too пarrow and the banks too steep and high for the maximum depth given Ьу these equatioпs to Ье developed without also induciпg an alaгmiпg hoгizoпtal exteпsioп.
Intersecting-jet spillways
The inteгsectioп of jets fгom ап oveгspi l l sectioп апd а mid-height spil lway сап Ье put to good effect to dissipate рагt of the eneгgy and reduce the exteпt of the scouг hole which woul d Ье cгeated Ьу а siпgle spil lway dischargiпg at the same total rate.
Thus, as shown Ьу Leпcastгe tests fог Alto Liпdoso dam i п Poгtugal, if а mid-height spi l lway discharging Q i is placed below а surface spillway d ischargiпg 0.5 Q1 so that their jets col l ide, the hole which гesu l ts from the total dischaгge ( 1 . 5 Q 1 ) need Ь е п о l arger, а п d сап еvеп Ь е s l ightly smaller, thaп that which would Ье scouгed Ьу the discharge (Q1 ) of the mid-height spillway аlопе [80, 8 1 \.
Howeveг, systematic hydraul ic model tests have yet to Ье dопе to appгeciate the ful l affect of the diffeгent paгameteгs (heads, veгt ical distaпce between spil lway outlets, width апd пumЬег of jets, di schaгge гatio of iпteгsecting jets, etc. ) .
А dгawback of th is arrangemeпt is that it сап сопsidегаЫу iпcrease the ri sks associated with spгays and mists (consequences on banks, transmissioп l iпes, switchyaгd, access гoads, etc. ).
Models
Phenomeпological models аге necessaгy in impoгtaпt cases, еvеп with widevalley sites, to give а betteг appгoximation of the shape апd size of а freely developing scour hole or to determiпe the protectioп гequiгed to inhibit development. The gгeatest difficulty гesides in ргорег representation o.f the erodaЬ!e materials. especially cohesive mateгials (non-homogeneous госk of vaгyiпg weatheгiпg and hardпess, which always contains at least some joi nts) апd some meaпs of veгification of the model, e.g., Ьу гeproduciпg а kпоwп past еvеп true to scale, would Ье пeeded.
7 7
The use of saпd or gravel of appropriate fiпeпess for а complete model of the river сhаппеl апd valley flaпks at а wide-valley site will yield the maximum fiпal size апd shape of the scour hole. As the aпgle of respoпse varies with particle size, scale effects are importaпt iп this type of iпvestigatioп. Maximum depth occurs at the poiпt of impact of the jets ; this is поt usually а cause for worry iп itself. With а bucket of appropriate geometry апd throw, retrogressive erosioп back from the impact poiпt will Ье moderate ; l ittle or по protectioп (cut-offs, large Ыocks) will Ье пecessary to protect the bucket ( especial ly for small flows wheп the jet does поt jump).
Back-curreпts however are frequeпtly the cause of excessive s ideways exteпsioп of а ful ly developed scour hole. The pattern of curreпts chaпge апd тау iпteпsify as the hole grows. Cut-offs, boulders, coпcrete Ыocks, coпcrete or rock spurs, апd various comЬiпatioпs thereof, сап stop such growth. Usefu l desigп iпformatioп оп such protectioп (depth of cut-off, weight of Ыocks, l iпiпg batter, etc.) сап Ье drawп from s uch tests.
Predictioпs оп the deve/opment of the scour hole in narrow-valley sites are much more hazardous. The р\оу of determiпiпg fiпal scour hole size with а moЬi le bed model ceases to apply, because the equil ibriu m profile would Ье reached after remova\ of а coпsideraЬ!e quaпtity of material, l iaЬ!e to completely uпdermiпe large structures апd produce serious Ыockages iп the river dowпstream. Coпditioпs where there is rock of exceptioпally good quality that would Ье hardly affected Ьу wear or damage despite proloпged attack Ьу dyпamic pressures, cavitatioп апd abrasioп are as rare as the large пatural waterfalls whose developmeпt appears to Ье arrested оп the humaп time scale. This has поt preveпted desigпers of ап iпcreasiпgly large пumber of spillways iп пarrow-valley sites from deliberately omittiпg апу protectioп iп the impact zопе. The implied explaпatioп is the low probaЬility of апу exteпded, heavy discharge whose iпevitaЫe coпsequeпce would Ье erosioп of the сhаппеl апd baпks ; they prefer to wait апd see, оп the argumeпt that erosioп wil l Ье slow eпough for suitaЫe protectioп to Ье provided before the situatioп becomes critical. This is а deliberate, " calculated " risk.
If the spi l lway is required to discharge high flows every year at ап uпprotected пarrow-valley site, substaпtial erosioп wil l Ье iпevitaЫe. А strikiпg example is the maiп spillway at Tarbela dam [ 1 9, 87, 88] which operates for several weeks every year at пearly 8 ООО m3/s (desigп capacity 1 4 ООО m3/s, head 1 30 m) ( Fig. 1 1 ) . The flip bucket throws the water iпto а пarrow valley орепiпg iпto the valley of а small tributary stream to the I пdus. Iп both valleys, the geology is complex with highly variaЫe rock, local ly quite hard but quite fiпely divided. No protectioп was provided except for а cut-off exteпdiпg sideways uпder the bucket. There has Ьееп coпsideraЫe erosioп represeпtiпg several mil l ioп cuЬic metres iп а few seasoпs s iпce first operatioп ( 1 974), seriously eпdaпgeriпg the bucket. M assive rol lcrete protectioп
79
plus а tied-back reinforced-concrete lining [68] have been found necessary, а total volume of the order of 700 ООО m3•
It is not easy to determine the protection пecessary at the start o f operation or becoming necessary \ ater. The designer does поt even have the imperfect tool of the mobile bed model, because the largest scour hole that сап Ье accepted is usually far less than the size of the fiпal hole iп \ oose material . Laboratories offer models built wholly or partly with materials of low cohesion (saпd and lime, saпd and с\ау, sand/chalk and cemeпt) c\aimed to represeпt rock and thus give information оп the exteпsion of erosion after а given time of operation. But the proper modelling of rocks of differeпt hardnesses, апd more importaпtly, of discoпtinuities, is unfortunately far from beiпg а scieпce. The prepoпderant role of discontiпuities iп such rocks (joints, fractures, faults) in the erosion process, because of their sensitivity to fluctuating dyпamic pressures, means that the results from а model material must Ье approached with cautioп, unless it has been possiЬ\e to calibrate the model after the first spilling has takeп place, to predict ultimate development, as was the case at Kariba.
Eпgineering judgement therefore p\ays а major part in designing protection in narrow-valley sites. The guiding idea is that the energy must Ье dissipated in а vo\ume of water half as large again as the volume that would Ье required iп а hydraulic jump for the same flowrate and head, and preferaЬ\y larger. Findiпg this volume may mean excavatiпg into the river chaпnel апd banks, as being preferaЬ\e to uncontrolled scour Ьу f\ow. l t is however desiraЬ\e that the scour hole should not Ье much wider thaп the nappe or jets, otherwise back-currents may form. А model with fixed sides will give valuaЬ\e information оп this point.
Means of' stahilizing scour hole growth are reinforced-coпcrete l iпings and/or layers of large rocks or artificial Ы ocks such as dolosse. Coпcrete l inings are much more resistaпt than rock to damage from fluctuating dynamic pressures because of their greater uпiformity. But in practice, they are nearly a\ways confined to protectiпg areas where they have а good foundation and backiпg such as rock of moderate to very good quality, otherwise they have to Ье prohibitively thick.
Blocks are only suitaЬ\e for horizonta\ or gently s\opping surfaces where they would поt Ье uпdercut, which greatly l imits their field of application.
Protection may Ье required for the bottom and/or sides. Typical of bottom protectioп is the coпcrete apron, usually below arch dams. But there are тапу examples of such aprons beiпg destroyed or seriously damaged right from the start, and one must \ook at the types and magпitudes of forces to which the apron is exposed.
8 1
I n the impact zone, the dynamic pressure may Ье equal to the total reservoir head if there is no cushion of water to soften the load (Fig. 1 2), but it decreases rapidly farther away so that the net effect is а concentrated load on the apron. Cola, Lencastre, H ausler, George and others (30, 56, 58, 79, 49] have attempted experimentally to fin d the relationships between steady pressure, head, discharge per unit width, jet angle with the apron, distance from point of impact and standing water depth. Some measurements from completed dams have also been reported. There is quite а high degree of scatter in the results although they all agree as to the importance of the depth of the water cushion on the apron. А consideraЫe depth may Ье needed to prevent the pressure reaching the bottom (Hausler reports 45 т for 1 00 m3 /s · т width for а height of 1 00 m).
This dynamic pressure, however high, should not raise an i nsurmountaЫe structural design proЫem ; the apron usually overlies rock and even if it has to span over а localized depression in the foundation, it is simply а matter of making it thicker, with more reinforcement.
Leaving aside the particular case of abrasive sediment load, the most dangerous factor is uplift pressure if а fiddure in the rock, or the rock/ concrete interface, communicates with the impact zone. This pressure propagates outwards without loss, and where it is not counterbalanced Ьу the fal ling j et above, it may act l ike а hydraulic jack сараЫе of l ifting the apron off its foundation and/or of displacing whole Ыocks of rock between natural joints.
This situation may Ье aggravated Ьу fluctuating dynamic pressures caused Ьу the high macroturbulence i n the impact zone [63]. Lencastre (79\ reports values 2.8 times higher than the average pressure in this zone if the layer of cushioning water is shallow (less than 1 1 .4 times the jet thickness), but both the average pressure and the fluctuating pressure became neg\igiЫe at the point of impact when the water depth was more than 1 5 times greater than the jet thickness, under his particular test conditions. M ore experimental research is needed. The influence of the height of the jet in particular does not seem to have been sufficiently i nvestigated.
Thus the most important type of damage to а plunge pool is similar to what was described to explain the destruction of the floors of hydraulic jump still ing basins. The risk of vibration of the slabs and fatigue in the concrete and reinforcement under these alternating stresses is also significant except if there is а sufficient depth of water to substantially reduce the magnitude of the pressure and the amplitude of the fluctuations.
Concrete aprons are set at а relatively shallow level, precisely because their purpose is to prevent scour down to а staЫe depth where the steady pressure and pressure fluctuations would cease to Ье significant. Even if there is no cushion of standing water, they can successfully withstand prolonged spilling provided they are not exposed to uplift pressures. The same constructional arrangements as for sti l l ing basin floors therefore apply : sealed joints, dowe\s and shear keys, and deep anchors
83
®
о � !1О � :;ю m </�CD
N
t ( 1
1 \ ' ' '
� '
® А.А.
84
Fig. 1 1
Tarbela dam service spillway (Pakistan). (А) Section along axis of service spillway chute
(1) Normal pool level (472. 75 m). (2) Guide wall. (3) Bucket lip. El 372. 10 т. ( 4) Protection wall. (5) Lower transverse drainage gal/ery. (6) Standpipe piezometers. (7) Longitudina/ drainage gallery. (8) Multiple point borehole extensometers. (9) Grout curtain.
(10) Ground /eve/ before spillway operationЕ/. 356.85 т.
(11) 7 sector gates 15.25 х 1 7.69 т.
(В) Service spi/lway plunge роо/ : Р/ап view.
85
(1) Spillway bucket. (2) Fully confining dyke 1 979-1980.
(3) Normal structura/ concrete. (4) Rockfill. (5) Drainage zones. (6) Rollcrete.
Та е/а dam (Pakista ). Spi ·ay.
86
Fig. 1 2
Surpression permanente sur u n tapis de protection sous une nappe deversante.
Permanent additional pressure over the apron surface below а free-fa/ling jet.
( \ ) Лh = Surpression permanente.
(2) Н = Epaisseur du coussin d'eau amortisseur.
(3) у = Distance horizontale depuis \е centre de l'impact.
D'apres R. Со\а : Si Н > 7,43 В. Лh max = 7,43 Vo'/2g х В/Н. Лh = Лh max х e- ''J'' · " .
87
(1) Лh = Additional permanent pressure. (2) Н = Thickness of water cushion.
(3) у = Horizontal distance from center of im-pact area.
After R. Со/а : lf Н > 7.43 В. Лh тах = 7.43 Vo
' /2g х В/Н.
Лh = Лh тах х е 43 "' /1
to prevent vibration (and resist any residual uplift pressure) [ 1 1 4]. The slab thickness and tonnage or reinforcement and anchor steel must Ье more substantial at and around the point of impact ; since this i nvolves only а relatively l imited area however, the extra cost is not prohibitive. Drainage can Ье provided at the rock/concrete interface provided the outlets are well away from the area of high pressure.
The extra protection provided Ьу а submerged apron is only significant if the water cushion is deep enough. As already mentioned, mode] tests and full-scale measurements reported to date are sti l l too incomplete to derive any general relationship between all the factors i nvolved. lt would appear that there is no appreciaЫe reduction i n the comЬined steady and fluctuating pressures if the depth of water is less than around 20 per cent of the head.
А small tai l water dam or weir is one way of obtaining а water cushion without the need for deep excavation (Vouglans dam, France) [ 1 1 , 27]. Care must Ье taken however not to reduce excessively the airspace underneath the nappe to prevent oscil lating flow conditions. I n the same way, consideration should Ье given to the risk of sediment bars forming downstream which might cause an undue rise in the tailwater level.
Side protection should Ье kept to the mш1mum necessary to prevent erosion Ьу the current emerging from the scour hole and the inevitaЫe back-currents. This means that the nappe or jets must not impinge on the banks where the depth of water is insufficient ; cuts i nto the bank are often necessary.
The turbulence associated with back-currents and the flow leaving the scour hole is consideraЬly less than in the impact zone. If the sides are far enough away, the fluctuating dynamic pressures on the side protection will Ье greatly attenuated. However, with vertical or battered concrete walls or l inings, а relatively low pressure on the rear face may Ье enough to cause them to topple. А l arge footing area and rock anchors will provide staЬility. А more difficult situation is where the side walls also shore up the bank. In any event, protection of the lower edge is the main issue. I f side protection is comЬined with а concrete apron, the scour hole is almost fully contained and the staЬility of the sides and apron comЬined into the same proЫem. The opposite situation is where the scour hole is allowed to develop freely during spil lage until the depth of water is sufficient for further deepening to proceed very slowly. I n this case, it is wise for the side protection built at the same time as the dam to Ье several metres deeper than the final scour hole depth, which means awkward trenching work. I n order to limit the depth of these walls, it is conceivaЫe to try to slow down the rate of deepening of the scour hole Ьу tipping in Ыocks once it has reached а certain depth ; but this is а much l ess rel iaЫe solution and may often involve insurmountaЫe practical proЫems.
89
Rock or concrete Ыocks for side protection are on\y feasiЬ\e in ttie case of wide-valley sites mentioned at ttie beginning of this section. They are used to stabil ize relatively gentle slopes inside the scour hole and up the banks to а comfortaЬ\e distance from the impact zone.
3.6.4. Model tests, Ref. [97]
I t has already been said that model tests are not always needed for the upstream and central parts of surface spil lways (frontal si l ls , side inlets or circular sil ls, approach channels, inlet shapes, shafts, chutes and tunnels) except where they differ wide\y from well-known types. When models are built this is generally taken as а good opportunity to examine conditions at the downstream end, for which the who\e surface spillway is commonly represented. This is why the following paragraphs are confined to the terminal structure.
The purely hydrau\ic design of hydraulic jump stilling basins only requires model tests if the configuration is significantly different from what has been done in the past. Otherwise it only serves to check the design.
If а mode\ is built, it must Ье run in such а way as to yield information оп the amplitude of the dynamic pressure tluctuations at various points on the wetted surfaces : \ower end of the chute, tloor and the foot of the side walls. Engineers now have the equipment for recording these tluctuations (piezoe\ectric cel\s) and for analysing the records (automatic determination of characteristic parameters : means, standard deviation, distribution, power spectra and so forth).
А scale model is essential on the other hand with flip buckets and nappes impinging into а scour hole.
Detailed design of bucket shapes that angle, divide, impact or spread the jet as required can only Ье donne Ьу experiment. The model will a]so measure loads on the bucket and determine the (usually small) discharge below which а standing wave will form on the bucket (this critical tlowrate will Ье higher when the discharge is increasing, lower when it is decreasing).
Estimating the final dimensions of the scour hole at open-valley sites Ьу means of а full mobile-bed model has already been mentioned. The chief objective of the test is then to check that retrogressive erosion is not likely to endanger the stability of the bucket of form bars across the river channel.
At narrow-valley sites, rea\istic determination of the staЬ\e scour hole means that the mode\ must reproduce the rock and its discontinuities, using а model material with the appropriate cohesion [67]. It has already been said that real progress in this area is still awaited. But the scale model is still а useful tool. Tests should begin with а bed materia\ offering no resistance against scour, and subsequent runs made with а slightly cohesive material . I f there are any well-defined barriers to erosion at the bottom, sides or upstream edge of the scour hole, they should Ье included in the model, preferaЬ\y after the first test with cohesionless bed material . Flow conditions at the fixed sides can then Ье investigated (whether or not
9 1
back-currents appear). Current ve\ocities, pressures and dynamic pressure fluctuations are measured at significant points and yield valuaЫe information on the erosion potential which the rock or lining must withstand. One can predict the l ikelihood of the u nprotected rock remaining undamaged (the probaЬi l ity is usually very low) or the utility of а given type of l i ning material . The depth of standing water and the sideways extension of the hole are two essential parameters for the test.
I f the hole is to Ье lined only at the sides (not at the bottom or downstream end), the concrete should extend downwards to а level just below the final depth of the scoured bed on the fixed-side model . The fixed sides representing the l ining must Ье i n р\асе before subjecting the movaЫe bed material to scour for the final depth to Ье а true estimate.
Measurement of velocities along the sides also he\ps in evaluating the risk of more or less coarse solids being entrained and resulting abrasion (especial\y if the bottom of the hole is not l ined). This risk can Ье control\ed Ьу building vertica\ linings that are not so exposed to abrasion, but the proЬ\em тау become more complex if these l inings also serve to shore up the banks ( deep anchors are nessary).
3.7. PROTECТION AGAINST ICE
lce can damage gate components, causing them to jam, and the only solution affording permanent emergency avail aЬi l ity is to heat vulneraЫe parts such as seals, drains, vents, grooves and guide rails .
Leakage down spillways can cause а bui ldup of ice which can divert flow or diminish the efficiency of ski-jumps and flip buckets. The freezing of mists and sprays has an equally adverse effect on transmission \ ines and building roofs, and it is for both these reasons that spillways sometimes discharge underground, as it is the case at M ica Creek and Portage M ountain in Canada [59, 1 50, 1 54].
J amming due to ice buildup can also Ье prevented Ьу heating the downstream side of the gate leaf itself (with insu\ation to prevent heat loss to air), or Ьу releasing compressed air from perforated tubes а certain distance beneath the water surface, the air bubЬ\es entraining warmer water to the surface near the gate leaf. Alternatively, warmer water taken from deeper in the reservoir, or generator cooli ng water, can Ье pumped to near the gates. Concrete piers can also Ье protected Ьу electrical\y heating their reinforcement bars [75].
93
4. ORIFICE SPILLWAYS
4.1 . COMPONENTS, LEADING CHARACTERISТICS, OBJECТIVES (*)
This Chapter considers orifice or submerged spil lways, ie, gated sluices and tunnels discharging downstream of the dam. Surface spillways with pressure Поw conditions at and for а short distance behind the inlet are discussed in sections 3.4 and 3 . 5 ; although there is no well-defined criterion for demarcation, the term orifice spillways is taken here to mean high-head outlets set at the bottom of the dam or at some substantial fraction of the dam height below water level, and whose discharge capacity represents а\1 or а substantial proportion of the total discharge capacity of the dam. N otwithstanding this definition, some remarks wil l also Ье made on submerged outlets of smaller size.
There are а variety of different arrangements of orifice spillways to suit discharge capacity, head, dam type, frequency of operation, downstream conditions and the availaЬ!e technology.
Recent progress, with gates in particular, now makes it possiЬ!e to have very large bottom sluices under very high heads for river Пооd discharge. There are some undisputed advantages. But one drawback may sometimes Ье that discharge capacity varies as the square root of the head, so that if the pattern of river Поw is not well documented, it may Ье necessary to provide а surface spillway as well for greater i nsurance against overtopping. The severe conditions under which large high-head spillways sometimes have to operate also demand а very high degree of reliaЬility and control from the hydro-mechanical equipment ; the designer must also know the properties of the materials used in the sluices to resist high velocity Пows.
Bottom outlets and high-head sluices may Ье required to serve any of the following services : (i) control and discharge surplus inflow, (ii) draw down the reservoir for dam maintenance, (ii i) control reservoir level at the critical time of first filling, (iv) draw down the reservoir rapidly in an emergency, (v) sediment П ushing, (vi) deliver water for project purposes, (vii) discharge river Поw during the construction of the dam, and (viii) Пооd prereleases:
(*) Ref. [ 1 1 , 2 1 , 27, 32, 1 36].
95
Compensation water releases, destined to maintain а specified flow in the downstream watercourse, are usually smal l and are provided through а special outlet.
А few examples taken from major dams will give а general idea of what has been done to date in this area.
The main purpose of sluices is discharging excess inflow l ike those at Kariba arch dam on the Zambia-Zimbabwe border where the six 8 . 5 m х 9 . 1 m s luices discharge 9 500 m3/s under а head of 33 m ( Fig. 1 3 ) . The spi l lway at Cabora Bassa [ 1 16, 1 19] in M ozambique (Fig. 1 4) consists basical\y of eight middle-\evel s luices with radial gates measuring 6 m х 7.8 m, for а total discharge of 1 3 1 00 m3/s under а 82 m head.
The five bottom outlets at Ohdo gravity dam in J apan measure 5 m х 5.6 m for а total discharge of 3 800 m3 /s under а 58 m head. They are designed for efficient reservoir management for flood control purposes. The main bottom outlet at М' J ara [ 1 1 7] earth dam in M orocco is designed for the same purpose but has а single 6.2 m х 6.6 m radia l gate сараЫе of 1 400 m3/s under а head of 73 m.
Sainte Croix [ 1 1 ] ( Fig. 1 5) arch dam in France has two 4.5 m х 4 m bottom sluices for а total of 1 1 00 m3 /s under а 72.7 m head. The B arthe [ 1 1 ] arch dam in France has two bottom outlets, measuring 3.5 m х 3 .3 m, сараЫе of twice 350 m3/s under а 59 m head. The functions are spi l lway, drawdown for maintenance, апd rapid drawdown in emergencies.
The seven 7 m square bottom outlets at Khashm Е\ Girba buttress dam in the Sudan provide the main spillway, some river floods being discharged after drawing down the reservoir to remove \arge quantities of silt [ 1 1 ] . M aximum capacity is 7 700 m3/s under а 32 m head (Fig. 1 6). The six 6 m х 1 0.5 m deep sluices at Roseires buttress dam in the Sudan ( Fig. 1 6), сараЫе of 7 500 m3 /s under а head of approximately 35 m, perform the same function [ 1 1 ] .
The main spil lway at Jupia buttress dam in Brazil [ 1 1 ] comprises 37 deep s lu ices measuring 1 0 m х 7.5 m and is сараЫе of 45 ООО m3/s under а head of 1 9 m (Fig. 1 6). This arrangement was adopted partly to hand]e the very large river flows during the construction period.
An example of large deep s luices for controlling the first-fil l ing operation is Manicouagan 5 ( Daniel Johnson) [37] multiple-arch dam i n C anada. Two sluices measuring 4.4 m х 3.5 m and controlled Ьу radia l gates were built to discharge 1 ООО m3/s under а 75 m head but also to Ье сараЫе of operating under а head of 1 50 m. These sluices were eventually plugged after the reservoir had been fi l l ed for the first time.
Compensation water in non-flood seasons is often provided through pressure pipes and valves, because of the finer control than with large gates. An example is the Grand Coulee gravity dam in the USA which has 40 pipes of 2.6 m diameter with а total capacity of 6 370 m3/s. Ring seal gates control the outflow. Another more recent examp\e is the Aldeadavi l la arch dam in Spain with two 2.5 m diameter hollow-jet valves discharging а tota] of 300 m3/s under а 1 20 m head ( Fig. 1 7 ) .
97
0 484 � G) 4 7..:;5,;,;· 8;,;_9=--
-""'""--llllil
О 10 20 m
Fig. 13
Kariba dam (Zamhia-Zimbabwe) : Section through orifice spillway.
98
(1) Six orifices, 8,5 х 9. 1 т2 each, controlled Ьуfixed wheel flat gate.
(2) Jet profil for reservoir at 475.80. (3) Jet profilfor reservoir at 494. 90. (4) Minimum tailwater level 382. 1 7 (283 m3/s). (5) Maximum tailwater level 404. 13 (9 627 m3/s). (6) Normal water level. (7) Minimum water level. (8) Scour hole.
Kariha dam (Zimhahive-ZamЬia).
99
�295
�231,00
Fig. 14
Cabora Bassa dam (MozamЬique) : Mid-depth orifice spillway : longitudinal section.
(/) 1 guard gate 6. 00 х 15.50 т 1. (2) 8 radial gates 6,0 х 7,8 т 1.
Total capacity = 13 100 m3/s. Maximum head = 85 т.
(3) Normal water level. (4) Maximum v.·ater level. (5) Minimum operating level.
1 00
4 8 1 .70
i
Fig. 1 5
Sainte-Croix dam (France) : Bottom spillway : longitudinal section.
1 0 1
( !) Two sector gates, 4 х 4.5 т 2 each, сараЬ/е together of 1 100 m3/s below reservoir EI. 477.
(2) Maintenance gate fixed wheel gate 4. 00 х 7.90 m2.
(3) Maximum reservoir Е/. (4) Minimum tailwater Е/. (5) Maximum tailwater Е/. (6) Splitters.
Fig. 16
Buttress dams - Typical section of bottom spillway.
(А) Minimum downstream water level
(В) Maximum downstream water level.
( 1 ) Kashm е/ Girba dam (Sudan) : 7 sluices with 7 х 7.30 т gates.
1 02
(2) Jupia dam (Brazil) : 37 sluices with 10 х 7.30 т gates.
® :;L,... 4 55. 50
+44о.о�
о 2 0 40 m
(3) Barrage de Roseires (Soudan) : 5 pertuis, vannes de 6 х 1 0,50 т. Roseires dam (Sudan) : 5 sluices with 6 х 10.50 т gates.
1 03
4.2. G EN ERAL CHARACTERISТICS OF SLUICES
I n discussing submerged outlets, а distinction must first Ье made between those provid i ng а very high discharge and those providing fine control capacity.
4.2.1 . Large bottom outlets
Large bottom outlets can Ье provided within the body of the concrete dam, or in а separate tunnel or conduit. In both cases, the control gate тау Ье at the upstream or downstream end, or at some intermediate point, as determined Ьу the general arrangement of the dam, flow conditions and arrangements at the discharge end. When the gate i s at the downstream end or at some i ntermediate position, the waterway ahead of it is usually steel-lined to protect the dam concrete against the ful l hydrostatic head i nside the sluice. Steel l ining is also placed for some distance behind the gate to improve cavitation resistance at very high flow velocities and wear resistance if the water carries а significant amount of sedime nt.
Broadly speaking, а bottom outlet requires two gates, the main control gate and а maintenance gate. Gate type as discussed below governs the general arrangement of the structure.
The sluice ahead of the control gate usually narrows down from the carefully shaped inlet to keep the pressure positive. The maintenance gate may Ье at any position along the in let section, but when it is at the downstream end stoplog grooves тау also Ье provided upstream of it.
The sluice behind the control gate may Ье steel-l ined for all or part of its length, to suit flow velocities, solid load if any, and whether or not there is а film of air entrained a long the walls. There are no clear rules and designs are usually based squarely on past ful l-size experience ; recommendations appear in section 4.4.
Large bottom outlets often have по debris screens (trashracks). Trees, branches, leaves, etc. may Ье sucked in, in large quantities and the easiest way of dealing with this proЫem is to make the sluices large enough not to jam and design the gates to open and close even with debris in the sluice. Where the slui ce is ! arge as compared with the size of the debris, no screen is used, and the i nlet shape is designed to prevent Ыockages.
1 05
When in doubt, however, а screen is provided ; it is very high and has the largest clear area compatiЬ\e with the size of the sluice. lt is designed structurally to resist forces t'rom total Ьlockage. Means must ot' course Ье provided for removing silt and debris if there is а serious risk of Ыockage (pressure jets, clear water feeding the seepage paths, see sections 4.5 and 4.6). For the usual bar spacing for trashracks, complete Ыockage is almost impossiЬ\e and some designers use an assumed and arbitrary differential head of 6 to 1 2 m .
4.2.2. Compensation outlets
Where fine control is required for smal\ Пows under high heads, various types of valve are employed, usual\y at the downstream ends ot' steel pipes. А sluice valve or gate may provide closure for maintenance purposes. Because of the small size of the Поw passage, they are usually screened. ln some cases they may also Ье comЬined with the large bottom outlet for discharging silt and debris . It must Ье remembered that there have been cases of sluices rendered unserviceaЫe through screen Ыockage and conduits Ь \ocked Ьу debris and silts.
4.3. GATES AND VALVES
Control gates for \arge bottom outlets may Ье radial or vertica\-lift type ; vertica1- l ift gates are usually used for maintenance gates.
For fine-control outlets, various types of sluice valve, usually at the lower end of the pipe, are used with а maintenance gate or valve ahead of them.
4.3 . 1 . Bottom outlet gates
The trend in the design of large bottom outlets is toward а more frequent use of radial gates for flow control. The great advantage where high heads are involved is that they do not need gate slots which disturb the flow and give rise to cavitation.
Radial gates have been tried successful!y at all heads up to about 1 00 т once efficient front seals had been developed. Gates with an offset at the bottom forcing the seal into place as the gate moves, have Ьееп tried at heads of more than 1 00 m, sometimes with offset aeration systems, but there is по conclusive evidence so far.
1 07
2 14.00 ---,
_у_
с
ш
Fig. 1 7
Aldeadavila dam (Spain) : Section through bottom outlets.
(/) Maximum water level. (2) Norma/ water level. (3) Scale metres. (4) 2 hollow-jet valves.
1О IS '20 rn
0 = 2.50 т ; н = 1 13.65 т. Q = 300 m3/s.
1 08
Aldeadi11·i/a dam fSpain). Spil/11•ay (2 800 m'/s).
109
Serre-Pon�oп dam (Francej. Spil/1 ay (1 m1/s).
1 10
Vertical-lift gates have been used as control gates under heads of up to 200 т (Mauvoisin in Switzerland), but they must remain rare instances and prolonged use at partial openings must not Ье а normal occurrence. This type of gate does of course have the advantage of taking up \ess space than а radial gate in sometimes congested areas.
The use of vertical-lift gates for flow control at partial openings requires special precautions under high heads. Up to about 20 m, one сап imagine that there would Ье few cavitation proЫems, but beyond this head, efficient operation requires special shapes for the s\ots and s luice to keep the flow at positive pressure and admit air into it [ 14\. There are instances of such control gates at heads of up to 60 т (Kariba spil lway, 33 m ; Р. К. Le Roux bottom outlet, 60 т (Fig. 1 8) ; Mica intermediate outlet, 60 т [84, 98, 99). Large vertical-lift gates сараЫе of operating at partial openings under heads in excess of 60 т are more rare. At Serre Pon�on, France, two outlets have 2 .60 т х 5 .60 т caterpillar gates and can each discharge 600 m3/s under 1 24 т head, but they are opened for only short time (annual check). At Colbun dam in Chile, two sliding gates measuring 2.5 т х 3.7 т are fitted in the diversion tunnel plugs and are сараЫе of 730 m3 /s total discharge under а head of 1 08 т uninterrupted operation for 324 days with flows between 50 and 690 m3 /s under heads up to 1 00 т. No cavitation proЫems were detected neither in the armor nor in the concrete.
1 1 1
о
0 RL 1 179.5 0 -=-у
...... -------
RL 1158,0 у
R L ll80,0
Fig. 1 8
Р. К. Le Roux dam (South Africa) : Bottom outlet and details of control slab gate.
(а) Layout of bottom outlet.
1 1 2
(1) Maintenance gate.
(2) Service gate.
(3) Slab gate 2.30 х 2.60 т '. (4) Splitter.
(5) Outlet house.
(6) Access gallery.
(7) Ventilation to bridge pier.
(8) Apron 3,0 т mini thickness.
(9) Maximum water level.
(10) Corresponding tailwater level.
(11) Apron le1·el 1082 to 1098.6 El.
960 140
о 2 4 б m
(Ь) Layout of slab gate.
1 1 3
(А) Sectional e/evation.
(В) Section А-А. (1) Transition from circular to rectangular sec
tion.
(2) Splitter.
(3) Slab gate.
(4) Gate housing.
(5) Hydraulic cylinder.
(6) Stainless steel lining.
An attempt has been made to reduce the hydraulic proЫems caused Ьу the gate slots at partial openings Ьу proposing solid « steel s lab » gates for Victoria dam in Sri Lanka (width 3 m , height 4. l m, head on si l l 94 m, thickness 270 mm), Alicura dam in Argentina (width 2.2 m, height 3 .2 m, head on s i l l 1 1 8 m, and thickness 260 mm) and Emosson dam in Switzerland (width 1 . 1 m, height 1 .8 m, head on si l \ 1 55 m, and thickness 1 20 mm).
Тhе downpu\I or uplift forces acting on а gate of given seal and housing geometry and the gate leaf vibrations should Ье analysed at various partial openings and operating conditions. M odel investigations can Ье used for these purposes as we\l as for the elimination of any upstream vorticies (cf. section 4.7.).
4.3.2. Maintenance gates for h igh-capacity bottom outlets
The maintenance gate on а bottom outlet must Ье аЫе to close against the full flow. It wil l usually Ье а vertical- l ift type (fixed roller, s l iding, caterpil lar) at varying distances ahead of the control gate as dictated Ьу sluice type. It may Ье at the upstream or downstream end of the gate inlet transition ; there may Ье one gate per sluice or one for several sluices depending on the risk factor.
The bulkhead type of gate which can Ье removed completely and stored above the maximum reservoir operating Jevel has а maintenance advantage.
Some maintenance gates can also Ье inspected in the operating chamber just above the s luice Ьу closing а horizontal Пар gate sealing the opening in the top of the s luice liner just be\ow the maintenance gate housing when the gate is in the ful ly open position. However, in this case the Пар gate itself is not accessiЫe for maintenance unless there is another bulkhead gate upstream.
4.3.3. Gates and valves for fine control
Conical hollow jet valves at the ends of pressure pipes are commonly used for fine control under high heads. At Glen Canyon dam (USA), four 2 .4 m diameter hol\ow jet va\ves can discharge 420 m1/s under а head of 1 62 m. At Botchac (Yugoslavia), two 2.8 m valves have а unit capacity of 1 28 m3/s. High pressure slide gates and jet Поw gates are a\so used for fine control .
The type of hol low jet valve that tends to Ье in most widespread use is the Howel l -Bunger va\ve with fixed cone and moving cylindrica\ s leeve. Construction
1 1 5
is quite simple апd it gives а good hol\ow jet without too much dispersioп if а hood is provided. I t does поt suffer from cavitatioп proЬ\ems provided it discharges iпto free space. The US Bureau of Reclamatioп cyliпdrica\ hollow jet valve a\so has excelleпt operatiцg characteristics. Iп this valve, the пееd\е moves to provide seatiпg.
For pipe mаiпtепапсе, butterfly valves, rotary (sphere) valves, s luice valves апd gates, etc. are provided. The spherical valve has а mаiпtепапсе seal.
4.4. СА VIТ А ТION, STEEL LINING, MAINTENANCE (See also Chapters 5 апd 6)
High-speed flow may iпvo\ve cavitatioп proЬ\ems depeпdiпg оп materials, shapes, surface fiпish апd aeratioп.
The rapid acce\eratioп of the flow uпder а high-head gate causes high flow ve\ocities aloпg the sides апd floor of the sluice with а very thiп bouпdary layer. It is usual to provide а steel lining for such а distaпce so as to permit growth of а bouпdary \ayer compatiЫe with the fiпish of the uпliпed coпcrete that fol\ows. But there is по souпd desigп rule апd the variety of differeпt \eпgths used is based squarely оп experieпce. Loпg liпiпgs are ofteп fouпd with high heads Ьеуопd 40 m. There has пevertheless Ьееп cavitatioп Ьеуопd the епd of the l iпiпg апd meaпs of admittiпg a ir have Ьееп provided as а remedy. For heads iп excess of 80 m, it would seem пecessary either to \iпе the who\e sluice behiпd the gate or admit air iпto the flow iп coпtact with the coпcrete surface at the poiпt where the liпiпg stops.
Radial coпtrol gates апd special s\ot shapes to avoid cavitatioп proЬ\ems have already Ьееп meпtioпed. Cavitatioп risks from mаiпtепапсе gate slots are less critica\ апd may Ье atteпuated Ьу the поw coпveпtioпal procedure of stream-liпiпg the slots [ 1 4\.
А staiпless-steel liпiпg will improve resistaпce agaiпst abrasioп from silt travelliпg through the bottom sluice. If wear does occur however, the l iпiпg is difficu\t to replace апd а resiп-based mortar, stee\ fibre coпcrete or other special liпiпg, may Ье preferred, as haviпg the advaпtage of епаЫiпg eroded surfaces to Ье repaired quickly. The proЫem of mаiпtепапсе апd access for mаiпtепапсе of s luices that haЬitually сапу sedimeпt must Ье giveп special atteпtioп.
Bottom sluices dischargiпg gravel must have special l iпiпgs, eg, graпite slabs, but the maximum ve\ocity must Ье kept quite \ow.
1 1 7
4.5. PROTECТION AGAINST FLOAТING DEBRIS, Ref. [57]
An abundance of floating debris can Ыосk inlet screens. The proЬ\em is critical if it can accumulate at the water's surface in front of the deep sluice апd it may Ье useful to provide а special surface discharge chute. Other meaпs of preveпtioп iпclude high screeпs апd keeping the deep sluice above river bottom. Stroпg trash booms will stop floatiпg debris but поt waterlogged material just below the surface. The example of Palagпedra dam (Switzerlaпd) is self explaпatory оп this matter : iп 1 978, а catastrophic flood led to а complete jamiпg of the spillway, the pressure tuппel iпtake апd both bottom outlets , with overflow оп the dam (cf. J . Bruschiп, S. Bauer, Р. Delley, G. Trucco : « The overtoppiпg of the Palagпedra dam » , Water Power, Jап. 1 982).
Ice which сап build up iп froпt of gates iп cold climates iп ofteп brokeп up Ьу ice breaker boat апd guided to special opeпings through which it is discharged.
If sti l l ореrаЬ\е iп frozeп coпditioпs (see sectioп 3 .7 . « Protectioп agaiпst ice » ), vertical lift or flap gates will allow discharge of ice iп а nappe up to 2 m thick. When the quaпtity of ice to Ье discharged increases too rapidly, as in case of ice jams, the gates are raised to their full height to completely clear the орепiпg. Iп this case, however, sti lling basins must Ье deeper and have no splitters or obstacles which could Ье damaged Ьу the Ь\ocks of ice [59, 1 50].
4.6. PROTECТION AGAINST SILTAТION
It is worth takiпg special precautions to protect sluices desigпed for discharging sedimeпt through density currents or flushing operatioпs at low reservoir level to protect them agaiпst becoming Ь\ocked.
I ril Emda dam iп Algeria апd Gebidem dam iп Switzerland are examples where special arraпgemeпts have Ьееп provided. At the first, the sluices tappiпg the deпsity curreпts are screeпed апd emerge at the sides of а larger sluice that is periodically орепеd to сlеап the side screeпs апd that can Ье fitted with stoplogs for mаiпtепапсе purposes.
At GeЬidem dam, where gravel has to Ье discharged, there was coпsidered to Ье а possiЬility of а deposit formiпg iп froпt of the sluice iпlets before flushiпg. А coпcrete screeп was therefore built in froпt of the iпlets to preveпt the deposit iпside the sluice buildiпg up to the roof. Clear water сап also Ье admitted behind the screeп to attack the toe of the deposit and counteract leakage under the gate, which might otherwise allow а plug of compact mud to form [85].
1 1 9
4.i . PROTECТION AGAINST VORТICIES
When the head above the entry of а partial\y opened deep surface spillway gate of mid-height spillway gate is insufficient, а vortex can Ье created, reducing flow and causing dangerous gate vibration and surge.
Vorticies can Ье e\iminated Ьу performing а scale-model study of special collars or Ыocks around piers, Ьу adding different types of flow-guiding surfaces (horizontal beams, vertical or subvertical training walls, gril les, cover slabs, etc.), Ьу injecting water from still zones [ 1 33] or Ьу boundary \ayer suction systems such as that studied for Piedro de Aguila dam in Argentina [24].
4.8. TERMINAL STRUCTURES AND ENERGY D ISSIPAТION (See Section 3 .6.)
There are two special aspects connected with the terminal discharge of flow from bottom outlets :
1 ) А highly concentrated jet discharge per unit width. 2) Operation under а high range of different heads.
For example, the maximum discharge per unit width from the Cabora Bassa spillway is approximately 270 m3/s.m under а head of 82 m ; at Khashm е\ Girba where the maximum discharge is 1 60 m3/s.m, the sluices must operate under а maximum head of some 3 5 m and а minimum head of а few metres only when the reservoir is drawn down in flood seasons to flush accumulated sediment.
At the Cabora Bassa, Kariba and Sainte-Croix arch dams, advantage is taken of the orifice-type flow to keep the terminal structures to а minimum, Ьу operating at low reservoir level and setting the orifices low enough for the impact zone in the river channel to Ье far enough away from the dam to prevent under-mining. The Morrow Point arch dam in the USA and Cambambe arch dam in Angola have the same type of orifice spillways but operating under а lower head, and large stil l ing basins had to Ье built.
At Khashm el Girba dam, the flip bucket is partially drowned at large flows. At Roseires dam where the deep sluices serve the same purpose as at Khashm el Girba, the stil ling basin is deeper but quite short, providing а hydraulic jump for energy dissipation at all discharges.
As with all spillways, the choice of free jet or stil ling basin with or without dispersion is closely tied with the overall dam design, topography, geology, operating conditions, etc.
1 2 1
5 . PARTICULAR PROBLEMSWITH HIGH-VELOCITY FLOWS
5.1. CAVIТAТION : GENERAL CONSIDERAТIONS, TYPICAL EXAMPLES (*)
Conditions conducive to cavitation аге aften combined in high-velocity flows : high turbulence associated with low ог moderate static pressure. Surface spillway chutes extending 40-50 т below reservoir level and their terminal flip buckets, and outlet channels from high-head spillways (see рага. 4.4) аге amongst the most highly exposed structures. Cavitation also occurs in the macroturbulent flow in hydraulic jump stil l ing basins and scour holes ; it is particularly intense at and behind the obstacles built into some types of basins.
Cavitation damage to structures designed to confine high-velocity flow is frequent with the higher dams and larger spil lways being built over recent decades. Taking only surface spi l lways as an example, such cavitation damage has occurred at :
Hoover dam, USA.
Glen Canyon dam, U S A [7, 1 55].
Karun, I ran, where the bottom half of the chute and bucket were seriously eroded on several occasions [ 1 62] (Fig. 1 ) .
- EI Infernil lo, Mexico, where cavitation scoured holes several metres deep in the bend joining the inclined shafts to the diversion tunnels.
- Keban, Turkey, where superficial damage occurred near the joints between chute slabs perpendicular to the flow [3].
Mica, Canada.
Bratsk, USSR.
Yellowtail, USA, where cavitation produced а hole тоге than 2 т deep downstream of the vertical bend in the tunnel spil lway [7, 20] (Fig. 1 9).
- Guri, Venezuela, where the first-phase spillway bucket was seriously damaged Ьу cavitation at the submerged lip [29].
(*) Ref. [40, 54, 74, 76, 1 14, 1 25, 1 56]
1 23
Е ф C\i со
( А ) _ _,..,..-
( D )
1 2 .95m dia.
5.ОЗma t heel
1 24
Е 1() (")
Е ...... ,...:
J- 9 .7 5m
Fig. 1 9 Evacuateur e n tunnel d u barrage d e Yellowtail (Etats-Unis) :
Disposition generale et details de l'aerateur.
Spil/way tunnel of Yellowtail dam (USA) : General layout and detai/s of aerator.
(А) Rainure d'aeration. (А) Air trough.
(В) Zone peu endommagee. (В) Area of minor damage. (С) Zone tres endommagee. (С) Area of major damage.
(О) Rainure en fer а cheval. (D) Horse-shoe trough.
(Е) Poche d'air. (Е) Air cavity.
(F) Detail de la rainure d'aeration le long de (F) Detail of air trough a/ong axis of invert. l'axe du radier.
1 25
5.2. CONTROL OF CAVIТAТION DAMAGE
5.2.1 . Surface finish, practical limitations, concrete ageing
The static pressure is generally low in free-surface high-velocity flows. When the streamlines are roughly parallel and near horizontal, the water depth is а measure of the static pressure, but it is less on а flat sloping chute (Pm = Уп х cos а, in which Уп is the thickness of the flow measured normal to the chute, and а is the angle the flow makes with the horizontal).
А convex floor reduces the static pressure, а concave floor increases it. Local variations are also caused Ьу surface irregularities to which are added the hydrodynamic fluctuations associated with turbulence. Subatmospheric pressures may appear momentarily near and, more importantly, downstream of irregularities.
The cavitation risk is significantly aggravated where high-velocity flows are in contact with the concrete surfaces, and the boundary layer is not fully developed [26]. This situation is unfortunately common with surface spillways ; the flow down the chute in а thick water sheet is still pseudopotential far down and centrifugal force at bends and flip buckets distorts the velocity profile unfavouraЬ!y.
Experiments in America (Ball [ 1 5, 16], Falvey [40]) and in Russia (Galperin [43, 44, 45, 46], Oskolov [ 104, 1 05], Rosanov [ 1 25], Semenkov [ 1 34, 135, 1 36]) have led to various equations for determining pressure and velocity conditions leading to cavitation or no cavitation for an isolated singularity on а chute floor, defined Ьу а given dimension or profi le. The tests would appear to have been made with а very attenuated boundary layer. The specifications for surface finish produced Ьу the Americans and Russians on this basis are in fact equivalent, the latter being slightly less stringent. For high velocities in excess of 30 m/s, the specifications are practically impossiЫe to follow in the present state of technology if they refer to as-cast concrete or screeded concrete. Finishing Ьу grinding is essential but this is laborious and has the disadvantage of weakening the surface because some of the superficial aggregate is inevitaЫy loosened. Contraction joint arises are particularly awkward to deal with. In any event, the finish quickly deteriorates with temperature fluctuations and freeze-thaw, even ignoring attack Ьу cavitation. Grinder-finish surfaces are particularly sensitive to weather damage.
I rregular surfaces can also Ье caused Ьу calcium carbonate concretions appearing after construction.
5.2.2. Surface treatment, linings and coatings (*)
I nadequate boundary layer where there are high-flow velocities а few millimetres from the wall justifies the measures described in section 5 .2 . 1 . for chutes and
(*) Ref. [ 1 , 53, 60, 73, 106].
1 2 7
buckets to surface spillways. Experience shows that there is practically no cavitation on а chute with а normal concrete finish if the boundary layer extends through the whole depth of flow and the favouraЬ\e effect of lower velocities near the walls outweighs the contrary effect of generalized turbulence. Unfortunately, the boundary layer will reach the water surface before dangerously high velocities develop (heads in excess of 50 m), only with thin sheets of water, ie, for small unit discharges.
Boundary layer growth is acce\erated Ьу surface roughness. This is why it is not necessary to demand а high standard of surface finish in the upstream part of the spillway.
А deliberately rough surface for some distance, with the roughness diminishing steadily from the top of the chute might Ье beneficial as regards cavitation control. This is worth exp\oring, but there would Ье practical difficulties of producing such а surface.
On а similar aspect, it is readily understood that а relatively long near-horizontal section of chute down to а point 30-40 т be\ow reservoir level is favoraЬ\e and should Ье used when the morphology of the site makes it feasiЬ\e. Simi\arly, long piers between inlets, possiЬ\y extending some way downstream, wil l contribute to boundary layer growth because of the interaction it promotes between friction on the floor and sides. The rooster tai\ of short piers, on the other hand, is favoraЬ\e due to aeration of the flow.
Special linings are used to improve surface finish and increase impact resistance against the imp\osion of the vapor-filled bubЬ\es produced Ьу cavitation. Both these objectives are met at one and the same time Ьу steel linings.
Steel linings have the disadvantage of being expensive and awkward to tie into the concrete. No extensive gap must Ье left between the steel and the underlying concrete because the dynamic fluctuating pressure may cause resonance so that the steel gradually separates further from the concrete and fatigue might ultimately tear the lining away. Steel l inings are general ly confined to small areas at risk. The most typical case is high-head spillways and bottom outlets plus the section of sluice immediately fol lowing. It is а wise precaution for the in\et to Ье а convergent shape so that the upstream pressure is always high enough to prevent cavitation, in which case, steel lining is only justified Ьу the need for the s luice to remain wtaertight and сараЬ\е of withstanding the inside pressure. The cavitation risk appears first just ahead (а few metres) of the orifice. Because of the rapid acceleration in the converging outlet, the velocity profile is unfavoraЬ\e over some distance (high velocities at the wetted perimeter) in the absence of any significant static pressure. Steel lining is inevitaЬ\e if there is а high probability of prolonged use at dangerous velocities in excess of 25 m/s, whether or not а vent is provided to aerate the flow near the orifice. There do not appear to Ье any precise rules derived from theory or experiment to determine the extension of the stee\ lining in the outlet сапа\, and this is still а matter of engineering judgement. The most prudent approach is to line the floor for а minimum distance of 50 Rн (Rн is the hydraulic radius of the orifice),
1 29
and the side walls over the whole wetted height for а distance of 1 5 Rн and up to тidheight for а distance of 30 Rн.
Since the object is to ассотраnу boundary layer to а point where the velocity against the walls Ьесотеs сотраtiЬ\е with the prescribed concrete surface finish, it is evident that the lining тust Ье longer for high orifice velocities and that increasing the surface roughness of the lining in the downstreaт direction would help to accelerate boundary layer growth. Tests shoul d Ье undertaken to arrive at а тоrе rational forтulation of lining length and better knowledge of the effect of surface roughness.
Another restricted use of stee\ as anticavitation lining is soтetiтes found on tlip buckets and in hydraulic juтps. The arrises or even the whole surface of sоте baftle Ыocks or obstacles are faced with steel to delay daтage. Although such protection is norтally subject to wear and replaceтent, progress is sti l l awaited to develop а convenient fastening systeт that would allow theт to Ье changed quickly.
Obtaining а sтooth joint between the steel and concrete is often а рrоЬ\ет where high tlow velocities are concerned. Teтperature effects, shrinkage and ageing тау cause the concrete edge to spall and adтit water that will tear away the steel plate. The lining тust Ье firтly tied-back and well drained to prevent uplift pressure, representing an appreciaЬ\e part of the velocity head, " jacking " the lining off the concrete.
Other types of protection besides steel have been developed over recent years. They are nearly always confined to repairs, because of their high price and \ack of knowledge about their long-terт perforтance. These are the ероху resins, fibre concrete and polyurethane resins*.
Ероху resins are appearing, which тау Ье brushed onto exposed concrete or iтpregnated to а depth of 1 -3 ст. Ероху тortars тау Ье used for repairs.
Painting and iтpregnating with ероху iтprove concrete surface finish and resistance, but тuch less than а steel lining. The iтpregnation process is laborious and difficult to apply successfully since the concrete first has to Ье dried to а depth of several centiтetres and teтperature requireтents are very stringent [ 1 0).
Ероху тortar is only suitaЬ\e for patching. Results are soтetiтes disappointing, especially when the repairs are exposed to а wide range of teтperatures - the
• Synthetic resins for facings of dams, ICOLD Bulletin 43, 1982.
1 3 1
edges of the patch spall off because of the different therтal expansion of the тortar and underlying concrete. The resulting rough surface is subject to cavitation.
Тhе use of steel fibre in concrete* is тotivated Ьу the experiтental fact that cavitation daтage is s lowed down Ьу increased tensile strength. Tensile strength in concrete is substantially iтproved Ьу steel fibres . The fibres also have the attractive property of тaking the concrete surface less brittle and preventing crazing. The fibres are 2-4 ст in length with а cross-sectional area of 0.05-0. 1 6 тт2• The latest repairs to the sti l ling basin to No. 3 tunnel at Tarbela ( 1 976- 1 977) included а 50 ст wear layer on the Пооr containing 80 kg/т3 steel fibre to iтprove cavitation and abrasion resistance [28].
There are sоте practical difficulties in тixing and placing fibre concrete because the fibres tend to bunch up and individual fibres poke out of sта\\ depressions forтing in the surface Ьу capi l larity as it is Пoated . The behaviour of the fibre concrete at Tarbe\a seeтs to have been satisfactory to date ( 1984) despite its prolonged exposure to siтultaneous uplift and cavitation froт the тacroturbulence in the hydraulic juтp. Fibre concrete has also been used to repair the cavitation daтage in the bottoт outlets of Libby and Dworshak daтs in the USA [ 1 20].
There is also another approach to protecting concrete against cavitation daтage that is quite different froт the idea of iтproving the surface tensile strength with hopes being pinned on the perforтance of thin elastic coatings that тight consideraЬly reduce the iтpact caused Ьу the iтplosion of the vapor-fi l led bubЬ\es . In fact, such ПехiЬ\е coatings are now being tried to сотЬаt concrete abrasion Ьу sediтent, but as iтpact is also а factor, а satisfactory answer against abrasion should also Ье largely suitaЬ\e against cavitation. The тost proтising ПехiЬ\е coating at the present tiтe appears to Ье polyurethane resin (see ехатрlе in section 5 .3).
5.2.3. Aeration Ref. [3 1 , 1 36]
Free air bubЬ\es in the Поw have а beneficial effect on cavitation. They oppose the developтent of subatтospheric pressures and Ьу increasing the coтpressiЬility of the water, consideraЬ\y reduce the force of the iтpact froт iтplosion of the vapor-fil led bubЬ\es.
Experiтents (Peterka [ 1 07, 1 08], Russel [ 1 28]) have deтonstrated that an eтulsion with sоте 8 % Ьу voluтe of air near the wetted surface eliтinates а\1 attack Ьу cavitation, even at very high Поw velocities in excess of 27 т/s.
• Fibre reinforced concrete, ICOLD Bulletin 40, 1982.
1 3 3
With free-surface flow, which is more critical thaп pressure flow as regards cavitatioп, there is а пatura\ aeratioп wheп the whole sectioп is turbuleпt to а sufficieпt degree for the surface teпsioп поt to preveпt exchaпges betweeп the liquid апd gaseous phases. As meпtioпed iп sectioп 5 .2.2, пеаr\у al l the turbuleпce is coпceпtrated iп the bouпdary layer at the top of а spillway chute, so that пatural aeratioп опlу begiпs опсе the bouпdary layer reaches the water surface.
Self-aeratioп is eпcouraged wheп the chute is loпg eпough to permit ful l developmeпt of the bouпdary \ayer ; i t has Ьееп used to reduce cavitatioп risk Ьу desigпiпg wide chutes so that, at пormal flowrates ( eg, 1 0-year flood) the flow is оп\ у about 1 m thick опсе velocities exceed 25-30 m/s (proposed spillway at Souapiti, Guiпea). The rooster tails produced Ьу piers with sharp edges at their trailiпg edge coпtribute sigпificaпtly to the self-aeratioп of the flow.
Iп а fully turbuleпt сhаппеl, the air сопtепt iп а giveп cross sectioп varies coпsideraЬ!y ; there is а substaпtial amouпt at the surface, but may Ье iпsigпificaпt at the bottom. If there are по obstacles or aeratioп devices, the mеап сопtепt will iпcrease gradually from upstream to dowпstream if the curreпt is accelerated because of the iпcreased turbuleпce.
Natura\ aeratioп of flows where the cavitatioп risk is high is rarely sufficieпt for adequate protectioп [25, 143, 1 60\. Ву the time it begiпs to Ье effective, the flow has ofteп reached а daпgerously high velocity апd there is stil l \ess thaп the miпimum 8 % of air at the bottom . But it is пevertheless wise to promote aeratioп. FavoraЫe factors are пaturally the same as those meпtioпed before iп соппесtiоп with bouпdary layer growth (sectioп 5.2.2).
Artificial aeration ( 1 7, 39, 1 1 8, 1 28, 1 34, 1 35, 147\ to preveпt cavitatioп iп free-surface flows leaviпg pressure sluices апd оп spillway chutes is comiпg iпto iпcreasiпg favor. The most widely used method is to create ап empty space uпder or at the side of the flow, iпto which atmospheric air is drawп. Iп practica\ terms, this is simply а step or slot with veпt pipes through the traiпiпg wal\s ; for example, the chute floor might Ье giveп а wide, deep slot апd/ or step, апd the side walls might have s\ots, steps, shafts or pipes, etc. (See Figs 20, 2 1 , 22, aeratioп devices of spillways for dams at Ust I l im апd Foz do Areia апd irrigatioп tuппels at Tarbela).
А ramp is ofteп added at the upstream edge of the step or slot оп the floor to deflect the bottom of the flow, which may jump from а few metres to 1 0 m апd more [33, 7 1 ).
Most of the eпtraiпed air is absorbed iп the iпteпse emulsificatioп at the bottom of the flow for some distaпce of the trailiпg edge. Performaпce, expressed esseпtial\y Ьу the ratio QA/QE (air rate over water rate), is goverпed Ьу а large пumber of factors (54, 1 09, 1 1 0, 1 1 1 , 1 1 2, 1 1 3 , 1 57\ of which the Froude пumber of the air-free flow appears to Ье ап importaпt опе, together with а dimeпsioпless
1 3 5
Е о <О l
SPIL LWA Y SEC TION
l
DETAIL I
Fig. 20
Ust Ilim dam spi//way (USSR) : General layout and details of aerators.
(А) Cavity beneath the jet.
(В) Ramp.
(С) Air vent outlets (2.2 т1 each).
(D) Dividing wall.
(Е) Air inlets (10 т1 each).
1 36
0EL
1 1 8. 50 27.
�---'-�=tl 1 . 50
Fig. 2 1
Foz do Areia dam spillway (Brazil) : General layout and details of aerators.
(а) Main dimensions of spillway.
1 3 7
(1) Aeration ramps.
(2) Joints.
(3) 4 gates 14.5 х 18.5 т1. Total capacity = 1 1 ООО m3/s.
(Ь) Aeration systeт of spillway.
Aerator 1 : d = 20 ст.
Aerator 2 : d = 15 ст.
Aerator 3 : d = 10 ст.
Basic air intake systeт.
1 38
Qa Air flow.
А Air outlet.
Foz do Areia dam (Brazil). Spil/11•ay - Cliute with aeratio11 ramps.
139
P L A N S E C T ION А - А
о 1 0 2 0 3 0 m
Fig. 22
Tarbela dam : lrrigation tunne/ п0 3 (Pakistan).
(а) Tunnel по. 3 air trough - Р/ап and section.
(1) Outlet gate chamber.
(2) Existing bulkhead slot.
1 40
(3) Concrete chute.
(4) Existing concrete walls.
(5) Air intake.
(6) Air shaft.
(7) Air trough.
* 3 45 2 5
®
1 1
Гn\ / 1 @ �l/ !
/ ( .r- JJ6,J5• 1 '
/ / , / /
о
Cross-section of air trough at end of steel liners of gate orifices 3 А and 3 В.
1 4 1
(1) Outlet gate chamber.
(2) Existing concrete.
(3) Existing bulkhead slot.
(4) Existing steel /iner.
(5) New deflector lip.
(6) Ероху bonding agent.
(7) Waterstop.
(8) Steel fiber concrete.
(9) 3 " х 3" chamfer.
(10) 1" х ]" chamfer.
2 m
parameter for the influence of the pressure differential across the water jet and the cavity length. The QA/QE = f(QE) relationship can Ье determined experimental\y for а given aeration arrangement. When using а scale model before the dam is built, scale effect is an important proЬ\em. Owing to the viscosity and surface tension of water, model results may underestimate the actual prototype performance. Viscous effects are due especially to the relatively thicker laminar sub-layer in Froude law models. Surface tension effects which inhiЬit the disruption of the water nappe seem to Ье consideraЬ\y reduced when the Weber number of the flow (We = pV2L/cr) is above 1 ООО ООО. То minimize scale effects, large-size sectional models may Ье used [ 1 091.
M easurements at dams (Tarbela irrigation tunnels [70), Foz do Areia surface spillway ( 1 1 0, 1 1 3]) reveal that the ratio QA/QE reduces as QE increases. For example, at Foz do Areia, QA/QE decreased from 0.7 to 0.07 for specific discharges varying from 1 0 to 1 1 0 m3/s/m ( 1 1 3). Care should Ье taken in the design for the aerators to attain satisfactory performance for the highest spillway discharge.
The efficiency of aeration devices is not governed solely Ьу the QA/QE ratio because the determining factor is the amount of air in the flow near the floor and the foot of the side wal\s (the upper part of side wal\s usually benefits from the abundant aeration produced Ьу the friction turbulence at the sides). There is usually more than the required 8 о/о where the water fal\s back to the floor but farther downstream, bubЬ\es tend to rise to the surface because of the diminishing pressure gradient and gradually escape. Air loss is greater at floor concavities (deflectors) because of higher hydrostatic pressures. After а certain distance, the air content may drop to below 8 %. On the 1 / 1 2 scale model of the Tarbela tunnels, nearly al\ the air entrained Ьу the bottom slot had reached the surface of the water 30 m below the slot ; however, the model results had shown consideraЬ\e scale effects. The question of the length of chute actually protected Ьу one aerator is still not well understood. Systematic observations on operational dams covering а wide range of discharges up to full capacity must Ье awaited.
Care is required in the construction of aeration devices, especially those in high velocity flows. Materials, surface finishes and structural precautions are the same as those recommended for parts exposed to dynamic uplift pressures and cavitation. Construction and contraction joints, and the edges of ramps and slots are subject to spalling and steel fibre concrete lining or small steel liner plates may Ье useful provided they are well tied-back into the concrete. Less strict finishing specifications are ассерtаЬ\е for the concrete boundaries where aeration is provided.
The emulsion produced Ьу high air contents causes а correlative increase in the cross section of flow. In а free-surface chute, the training wal\s have to Ье higher ;
1 43
with а tunnel spillway, the increased cross-sectional area required to prevent the tunnel running ful l may Ье costly. At Foz do Areia spillway, for а discharge of 8 500 m3/s and velocities of the order of 43 m/s bulking reached about 80 %. For these reasons, one may wonder whether research should not Ье directed towards an aeration system consisting of small devices only some dozen metres apart along the chute, so as to maintain the 8 % air content necessary on the bottom without producing too much emulsion in the rest of the flow.
5.3. ABRASION IN SPILLW А YS AND ВОТТОМ OUTLETS SPECIAL SURFACES
5.3. 1 . General Ref. ( 1 06)
Bed load and suspended load in а fast-flowing current can cause rapid wear of the sluice. In many cases, wear is inevitaЫe because there are as yet no known materials that will resist for а long time, are easy to use and reasonaЫy priced.
Fine clays and silt in suspension do not generally raise any serious proЬ!ems. This is not so however with sands in suspension or rolled along the bottom when they include hard, angular grains, or gravels and реЬЫеs of which а significant proportion, if not all, is commonly abrasive.
The proЫem will appear sooner or later with all bottom outlets and deep spillway sluices. With surface spillways, the proЫem will not usually arise until many years in the future, except in the case of small reservoirs that fil l up with sediment in the first few seasons' operation.
Although of the same fundamental type and often comЬined, а distinction will Ье made between whether the bed load or suspended load is the more significant factor.
5.3.2. Abrasion from bed load
Worrying situations are found at dams built on fast-flowing mountain rivers. Once there has been some siltation in the reservoir, or sometimes right from the first fil l ing, the bottom outlets and/or spillways have to discharge very large quantities of solids.
The parts most exposed to abrasion are floors and the foot of the side walls. Design arrangements recommended to control or slow down the proЫem are :
keep passages as short as possiЫe (arch dams are best in this respect) ;
- use concrete aggregate with high resistance to friction wear ( eg, quartzite sand and gravel for instance) ;
1 45
- keep passages free from singularities (bends, piers, converging or diverging transitions) ;
- provide critical areas with special surface finishes. Linings and coatings are the same as for cavitation protection (see section 5 .2 .2).
Experience has shown that such protection reduces the frequency and extent of periodical repairs but it is never а permanent solution.
An example is the spillway at Ku Kuan dam (Fig. 2) built in 1 959- 1 962 on the Та Xia Chi mountain river in Taiwan, which carries enormous quantities of quartzite sand and реЬЫеs with each flood caused Ьу the typhoons. In the first two years the reservoir silted up to the lips of the spillways at midheight up the dam (which is an arch dam 85 m high). Since then, all river floods with their ful l solid load have been discharged through the four 9 х 6.6 m2 sluices which have stainless-steel l inings оп the floors and the lower third of the side walls. The control gates are fixed roller types on the dam downstream face with the roller track set well back from the edges of the sluices to keep them clear of sedimeпt impact. The jets fal l iпto а basiп liпed on the bottom апd sides with reinforced concrete 80 m long tied-back into the rock Ьу means of steel саЫе. Drains at the rock/ concrete interface discharge near the downstream edge. Average concrete thickness on the basin floor is 3 . 5 m, increased to 5 m over the small area where the jets fall .
Ku Кuап dam was commissioned in 1 962 and the sluice steel linings have stood up well to wear. The stil l ing basin floor has required periodic repairs in the impact area, the latest ( 1 982) after а period of eight years' operation when а total of 550 hm3 of water had been spilled, mostly with the reservoir at its top level (80 m head). It is estimated that at least 550 ООО tonпes of bed load ( \ kg per cuЬic metre of water) fel l onto the concrete in this time. Almost all the erosion was confiпed to the impact zone where а pit up to 4.3 m deep and 700 m3 in size was formed. There was no evidence of any significant contribution from other factors such as uplift or cavitation. Downstream of the i mpact zone, there was abrasion wear оп the floor and sides but it is опlу superficial and very minor compared to the situation in the impact zone.
А layer of flexiЫe Ьituminous concrete laid on the bottom of the basin at the outset was quickly torn away and was пever reпewed because the proЫem of providing а bond with the underlyiпg coпcrete was found to Ье iпsurmoпtaЫe.
As а general comment the construction of ап abrasioп and cavitation resistant lining done to date оп а worldwide basis has rarely been successful апd stil l requires а minimum maintenance.
5.3.3. Abrasion from suspended load
Abrasion caused Ьу suspended load can occur аlопе or in comЬination with damage from bed load. The proЫem is usually associated with fine to very-fine sand or even silt containing а high proportion of angular quartz grains.
1 4 7
Example 1 : Yellow River, Sanmenxia reservoir
Drainage area : 684 ООО km2 Mean annual inflow : 42.3 km3 Suspended load : 1 .6 Gt/year Suspended load in July to October : 1 .38 Gt Mean annual solid load : 38 kg/m3 Mean solid load, July to October : 57 kg/m3 О 1 00 = 0.30 mm ; О 50 = 0.03 mm ; О 1 5 = 0.005 mm. 90 % including minus О .О 1 mm fraction is angular quartz grains.
Example 2 : Atbara (а Nile tributary), Khashm el Girba
Mean annual inflow : 1 2 km3 Solid load : 90 Gt/year Mean annual solid load : 7.5 kg/m3 sand 52 %, silt 28 %, clay 20 % sand Dso = 0.2 mm, 090 = 0.35 mm.
The most frequent risk of abrasion from suspended load occurs in the same structures as for bed-load damage : bottom outlet and high-head spillways where flow velocities are usually high and where the solid load content is much higher than the average over the whole depth of the reservoir (more than douЬ\e the average at Sanmenxia dam on the Yellow River), which greatly complicates the proЬ\em of surface wear.
Chinese experience at Sanmenxia has shown that wear on the floor, expressed as the mean thickness е removed, is given Ьу an expression of the type
k V3 t т е = ---
Rн
in which V is the mean flow velocity, t is the sediment content, Т is the period of operation for а given V and t, Rн is the hydraulic radius, and k is а coefficient depending on Dso and sediment shape.
An important parameter that is ignored in the above relationship which refers to а straight sluice, is curvature of the streamlines. Wear is much more rapid on concave surfaces. If the flow separates from the surface and then comes in contact with it again, there will Ье consideraЬ\e wear at the point of impact. This is often aggravated Ьу cavitation once the wetted surface has been roughened in this way.
The formulation for abrasion wear is still far from perfect. Observations at Sanmenxia show that (i) wear of ordinary concrete surfaces and steel l inings is insignificant at velocities of less than 1 О mls, despite the high sediment contents mentioned above, (ii) wear is still negligiЬ\e up to 1 2 m/s for concrete surfaces containing а high percentage of quartzite, but begins to appear on steel linings, and (iii) there are materials offering reasonaЬ\e duraЬility for velocities of less than 25 m/s (see section 5.3 .4). At higher velocities there is at present no way of providing effective protection with significant amounts of abrasive sediment (beyond а few kg/m3) in the f\ow.
1 49
Iп order to keep таiпtепапсе to а тiпiтuт, structures carryiпg suspeпded loads тust :
1 ) Ье short ; 2) Ье large eпough for тost of their -leпgth to keep flow velocity to arouпd
1 0 т!s ; 3) produce high velocities опlу over the last few тetres before the outlet where
replaceaЫe or repairaЫe liпiпgs сап Ье provided ; 4) поt al\ow these high velocities to exceed 25 т!s ;5) discharge back iпto the river сhаппеl iттediately behiпd the outlet, апd6) Ье easily апd quickly accessiЫe for iпspectioп.These recoттeпdatioпs are siтilar to those for cavitatioп coпtrol. The secoпd
опе тakes а pressure coпduit iпevitaЫe ; the fourth теапs that bottoт outlets or spillways тust поt Ье operated at heads of тоrе thaп 30 т.
5.3.4. Surface protection
Surface protectioп agaiпst abrasioп is the sате as for cavitatioп. Steel linings are still the тost widely used. Froт evideпce at Sаптепхiа iп Chiпa, а coпcrete with quartzite aggregate тight Ье preferaЫe, but there are по other cases to corroborate the evideпce that it perforтs better.
The bottoт surfaces of the bottoт outlet at Khashт el Girba ( Fig. 1 6) were protected with 1 2 тт of steel which abrasioп wore right through after 1 2 years' operatioп. lt was calculated that 8 .5 Mt of abrasive saпd had passed over each тetre of width, the тахiтит velocity beiпg 2 1 .5 т!s.
The staiпless-steel liпiпg to the spillway sluices at Ku Кuап dат have resisted extreтely well after dischargiпg тоrе thaп 1 .5 kт3 of heavily \аdеп water over а period of тоrе thaп tweпty years.
The таiп difficulty with steel liпiпgs is the fixiпg of rер\асетепt plates iпto the coпcrete. There is still по satisfactory solutioп for quick, re\iaЫe repairs, but it will соте eveпtually.
Нigh cement contents (600 kg per cuЬic тetre of coпcrete) with the additioп of corundum апd granite setts have Ьееп used with varyiпg aтouпts of success for critica\ surfaces at low-head scheтes (Сотраgпiе Natioпa:le ·du Rhбпе) [92). At Sаптепхiа, the coпvergeпt outlets froт the bottoт sluices wЪere velocities сап reach 25 т/s had their floors protected with vitrified paviпg (kпоwп as artificial diabase slabs) 2 ст thick апd bedded оп тortar. These sluices have Ьееп operatiпg coпtiпuously for 4 тouths per year with solid loads raпgiпg froт 63 to 1 83 kg/т3 (with exceptioпal peaks of 476 kg/т3). There has поt Ьееп тоrе thaп 1 тт surface wear after 5 years.
1 5 1
As for cavitatioп, thiп-film finishes with neat ероху resins апd ероху mortar are used occasioпally for patching eroded coпcrete. Temperature Пuctuatioпs shorten the l ifespaп significaпtly. Recently, new ероху liniпgs have been developed which are поt sensitive to temperature Пuctuatioпs (with ап ехрапsiоп coefficient close to that of coпcrete) апd quite promisiпg. Pitch-epoxy-corundum mortars have Ьееп tried at the Villeпeuve апd Peage-du-Roussilloп dams оп the Rhбпе, апd at the Viщ:a dam in the Pyrenees, iп layers 20 тт thick ; they are more resistant to abrasioп thaп epoxy-resiп mortars because of their greater ПexiЬility.
The most strikiпg inпovatioп iп receпt years was the substitution of а ПехiЫе polyurethane resin for the steel l iпing at Khashm el Girba daпi already meпtioпed. Iп а test developed Ьу the Compagпie Nationale du Rhбne, the performance of the polyurethane resin was remarkaЬ!y better than other materials. The test coпsisted of shootiпg а jet of water and solid silica опtо the sheet of material tested at ап aпgle of 45 degrees and а pressure of 0.25 МРа. The sample is in а pan filled with water. The staпdard test lasts for 75 miпutes, after which the volume of the pit eroded from the surface is measured. Measured volumes from the test were as fol lows, using а glass plate as reference, with а coefficieпt of 1 :
Normal coпcrete, 350 kg cemeпt per cu.m coпcrete Coпcrete with 600 kg cemeпt per cu.m coпcrete + coruпdum Epoxy-resin mortar Neat ероху Good graпite Pitch-epoxy-coruпdum mortar Neat, smooth polyurethaпe resiп Steel Cast iroп Epoxy-urethaпe-haematite
3.50 to 4.00 0.90 to 1 .00 0.80 to 0.90 0.22 to 0.24 0.55 to 0.75 0.50 to 0.60 0. 1 4 to 0. 1 80.04 to 0.05 0.02 to 0.03 О.Об to 0.08
The work at Khashm el Girba after removiпg the steel liner coпsists of sandЫasting the concrete and primiпg the surface with ап ероху mortar, spreadiпg а 1 4 mm layer of polyurethaпe resiп mortar and finishiпg with ап 8 mm wear coat of neat polyurethaпe resiп.
The surface has stood up well to wear but uпfortuпately submerged ЬаоЬаЬ and other timber has kпocked off patches and repeated maiпteпance has been found пecessary.
5.4. NIТROGEN RELEASE IN SPILLWAY AND OUTLET TAILWATERS
Betweeп 1 968 апd 1 970 the effect of high пitrogeп saturatioп iп waters below spillways was recogпized. This has disastrous effects оп fish l ife dowпstream апd
1 53
particularly affects salmon ( 1 38]. In 1 970, it was reckoned that nearly 90 % of seawardbound salmon in the Snake River in the United States died of the effects of the supersaturation in nitrogen of waters discharged from spillways.
This so-called nitrogen oversaturation (which is really an overabundance of dissolved nitrogen and oxygen) is а new determining factor in the studies of water qua\ity for hydrau\ic projects on rivers where fish play an important ro\e. Experience has shown that supersaturation is directly related to the rate of discharge, becoming asymptotic at some \evel as the intensity of spill is increased.
То he\p overcome the proЫem, it is necessary to restrict discharge and increase the storage capacity upstream, taking care not to flood the spawning grounds upstream. This is not in itself а so\ution however.
TurЬined waters do not have а higher content in dissolved gases, but this is not а standard so\ution since power demand is often reduced during the flood period. For this reason, the US Corps of Engineers has studied the possiЬility of inefficient power generation, producing \ittle power for only а slight reduction in discharge : this involves p\acing а multi-orifice bulkhead in the stop\og housing upstream of the power intake and ajusting the runner Ьlades (Kaplan type) to obtain minimum efficiency. It remains to Ье proven Ьу long-term testing that the runner does not suffer from such treatment.
Another method is to place а deflector at some point at the base of the spillway chute in order to direct а moderate flow (up to 425 m3 /s for widths of 1 5 m at the Lower Granite dam, United States) to the tailwater surface rather than the bottom.
Lastly, the ultimate solution to the proЫem is to portage migrating fish around areas where river water has а high nitrogen content. But here too, it remains to Ье proven that the use of specially-equipped trucks will not disturb migratory instincts and prevent adult fish returning to their spawning grounds.
1 5 5
6. MAI NTENANCE AND REPAIR
6.1 . MAINTENANCE
The maintenance of the spillway civil works raises no particuJar proЫems as compared with the rest of the dam and appurtenant structures. The surface finish required on parts liaЫe to Ье eroded Ьу cavitation do however merit more attention than other concrete surfaces. Flaking or spalling from weather or accidental knocks from tools, scaffolding, stones, etc. must Ье repaired before the spillway is reopened. А special watch must Ье made of surfaces finished Ьу grinding because they are more fragile. Deposits and hard concretions (carbonates) must Ье removed.
Gate maintenance is obviously of the greatest importance. The best solution is to operate gates for maintenance purposes under full or average head, but in many cases gates permanently hold back water, and opening them for а complete maintenance operation would involve unacceptaЫe losses of reguJated inflow and cause an unwanted artificial flood downstream. Some arrangement of stopJogs or guard gate permitting periodic rnaintenance of the gate and built-in parts is therefore essential.
In attempting to reduce maintenance for corrosion, the trend is for eliminating саЫе winches, especially where the саЫеs are permanently under water, the preference being for solid members l ike link rods or sprocket chains. Stainless steel for roller paths and seal mating surfaces in continuous contact with water is still the general rule.
Gated spillway trials at least at partial opening, depending upon the ассерtаЫе flow downstream, have to Ье done whenever feasiЫe. Regular safety inspection (gates and operating mechanisms, chute or tunnel, bucket, plunge pool or stil l ing basin, downstrearn scour) have to Ье made regularly (see ICOLD Bulletin 29 and 49).
6.2. REPAIRS
The most frequent type of damage calling for urgent and often extensive repairs is that caused Ьу cavitation, abrasion and energy dissipation. The parts most exposed to one or more of these factors have been discussed in earlier chapters, together with how the damage is caused and means of attenuating its effects. But whatever precautions are taken, the probability of damage occurring is never zero, especially if large flows are discharged frequently. The only hope is that damage from an isolated flood event wi ll not reach disastrous proportions and there will Ье time and
1 5 7
resources for repairiпg it before the пехt tlood. This attitude implies that suitaЬ\e provisioпs for repair are iпcorporated at the desigп stage.
Regardiпg cavitatioп erosioп оп chutes апd buckets, а wise precautioп iп situatioпs of high risk is to split the total discharge capacity up amoпg differeпt outlets апd provide the dam operator with proper р\апt for rapid repairs. As а geпeral rule, chutes апd buckets are directly accessiЬ\e wheп поt iп use because they are well above tailwater level, which greatly facilitates iпspectioп апd repair.
Abrasioп raises similar proЬ\ems, the solutioп to which is a\so easier if the relevaпt areas are easily accessiЬ\e.
The most critical situatioпs are caused Ьу damage from eпergy dissipatioп dyпamic uplift pressures, abrasioп апd cavitatioп. Hydraulic jump stilliпg basiпs are particularly worryiпg because their positioп below tailwater level makes iпspectioп апd repair difficult. Uпless the river is completely dry they сап оп\у Ье iпspected апd repaired after beiпg isolated апd dewatered. Their widths апd depths mеап that the cofferdam or stoplogs are ofteп very large апd slow to iпsta\\ . H ere agaiп, the desigпer would Ье well advised to split the discharge up betweeп several iпdepeпdeпt basiпs апd, as far as possiЬ\e, provide piers апd а gaпtry оп or just dowпstream of the outlet sill where the stoplogs, ready оп site, сап Ье easily iпserted.
Extra safety is sometimes iпcorporated Ьу dissipatiпg the eпergy iп two stages (eg, Maпgla dam, Pakistaп) with two stilliпg basiпs, the first beiпg set just above tailwater \evel. I t is more likely to Ье damaged Ьу the higher head, but it is also more easily accessiЬ\e for repair.
The difficulty of iпspectiпg апd rерашпg hydraulic-jump stil liпg basiпs is sometimes а determiпiпg factor iп favor of а scour hole. Iп the best cases the hole will develop freely апd reach а пatural staЬ\e profile without much eпcouragemeпt. Protectioп, if апу, is geпerally at the sides, exteпdiпg dowп to well below the predicted scoured depth so that there should пever Ье апу пееd to dewater the hole. But, if there is ап арrоп to limit the depth of erosioп, its much greater vulпeraЬility implies that steps must Ье takeп to facilitate iпspectioп апd repair except if it is dry or just awash wheп the spi\lway is поt operatiпg. Such measures iпclude loпgitudiпal compartmeпtalizatioп, а tailwater dam, or stoplog piers.
The repair works at Kariba are worth meпtioпiпg. Uпderwater iпspectioп Ьу techпically qualified divers has Ьееп carried out iп the 60 т deep scour hole siпce 1 962, where successful coпcrete repairs to the upstream sides have Ьееп completed
159
on а regular organized basis, using divers to clean, shutter and place rockfill underwater, comЬined with dril ling and grouting from floating platforms and depth charging to deter crocodiles. The second powerstation recently completed on the opposite bank of the Zambezi has substantially reduced spillage which was 25 km3 per year on average up to 1 98 1 and the attendant repairs. The plunge pool has remained practically unchanged since that time.
1 6 1
APPENDIX 1 REFERENCES
1 . АвЕLЕУ А. et а! . . 1 97 1 : " I nvestigations of Relative Cavitation Resistance of Material and Protective Coatings and Development of Measures against Cavitation Erosion of Hydraulic Structure Elements ", Proceedings 1 4th I A H R Congress, Vol . 5 , Paris.
2 . Анмло 1 . , С н 1sнп М . 1 . , RлsнEED М . А., 1 979 : " Tarbela Left Bank lrrigation Tunnel = Some Special Features of Design, Construction and Operation ", 1 3th ICOLD, Q. 50 - R. 63, Yol. 5, New-Delhi.
3. Aкsov S., Eтн EM BABAOGLUS S. , 1 979 : " Cavitation Damage at the Discharge Channels of Keban Dam ", 1 3th ICO LD, Q. 50 - R. 2 1 , Уо\. 3, New-Delhi .
4. ALEXANDER G. N. , 1 973 : « Estimation of the 1 0 ООО Year Flood » , l l th ICO LD, Q. 4 1 - R. 70, Yol . 2, Madrid.
5 . ALVARES R I B E I RO А., PINTO ол S I LVA F. et а! . . 1 967 : " Erosion in Concrete and Rock due to Spil lway Discharges " , 9th ICO LD, Q. 33 - R. 1 8, Vol . 2 , I stanbul.
6. ARMSTRONG R. С., 1 979 : " Corrosion of Outlet Tunnel Concrete on Two Muskingum River Basin Projects USA ", 1 3th ICO LD, Q. 50, Discussion 36, р. 570, Yol. 5, New-Delhi .
7. ARTH U R Н . G. , JдBARA М . А., 1 967 : " ProЫems involved in Operation and Maintenance of Spillways and Outlets at Bureau of Reclamation Dams ", 9th ICOLD, Q. 33 - R. 5, Уо\. 2, I stanbul.
8 . ASCE, 1 973 : " Re-evaluation of the Adequacy of Spillways of Existing Dams ", Ат. Soc. of Civ. Engrs., Уо\. 99, No. Н У.2 ( February).
9. ASCE Task Force, 1 973 : " H igh Head Gates and Valves in the United States » ,Ат. Soc. of Civil Engrs., vol. 99, No. Н У. 1 О.
1 0. ASI LOMAR Conference, 1 979 : " I mproved H ydrologic Forecasting - Why and ' ноw », Ат. Soc. of Civil Engrs. (March).
1 1 . AuвE RT et а!. (Comite Fraщ;ais des Grands Barrages), 1 973 :« Les ouvrages d'evacuation definitifs des barrages », 1 1 е CIG В, Q. 4 1 - R. 35, vol . 2, Madrid.
1 2. Auв1N L., LEFEBVRE D., SтOLAN А., 1 979 : « Les evacuateurs de crues des amenagements hydroelectriques LG 2 et LG 1 du complexe La Grande » , 1 3е CIGB, Q. 50 - R. 8, vol. 3, New-Delhi.
1 3 . Влек Р. А., F R EY J. Р., JOHNSON G., 1 973 : " Р. К. Le Roux Dam = Spillway Design and Energy Dissipation ", l l th ICOLD, Q. 4 1 - R. 76, vol . 2, Madrid.
1 4. BALL J . W., 1 959 : " Hydraulic Characteristics of Gate Slots " , Ат. Soc. of Civil Engrs . . No. Н У . 1 О (October).
1 5 . BALL J. W., 1 963 : " Construction Finishes and High Yelocity Flow ", Ат. Soc. of Civil Engrs . . No. СО 2 (September).
1 6. BALL J. W., 1 976 : " Cavitation from Surface I rregularities in H igh Yelocity " , Proceedings Ат. Soc. o.f Civil Engrs . . vol . 1 02, No. Н У.9 (September), Discussion Ьу ARN DT, vol. 1 03 , No. Н У.4 (April 1 977).
1 62
1 7 . BEICHLEY G. L. , 1 975 : " Cavitation Control Ьу Aeration of High Velocity Jets " , Proceedings Ат. Soc. of Civil Engrs., vol. 1 0 1 (July) .
1 8 . BELLIER J . , BAUZIL М ., 8ARGE, ВLЛNСНЕТ et а! . . 1 967 : « Evacuateurs de crues " longs " » , 9е CIGB, Q. 33 - R. 40, vol. 2, I stanbul.
1 9 . B I NGER W. У., 1 978 : " Tarbela : Plans, ProЫems and Success " , Water Рои•ег and Dат Construction (July) .
20. BORDEN R. С. , COLGAП: D., LEGAS J. , S ELANDER с . Е., 1 97 1 : " Documentationof Operation, Damage, Repair and Testing of Yellowtail Dam Spillway ".REC- ERC-7 1 -23, US Bureau of Reclamation, М ау 1 97 1 .
2 1 . BOURG I N , Ruпт, VoR M E R I N G E R et а! . . 1 967, « Considerations sur la conception d'ensemЫe des ouvrages d'evacuation provisoires et definitifs des barrages », 9е CIGB, Q. 33 - R. 27, vol. 2, I stanbul.
22. BoW E RS С. Е., Тsл1 F. У . . 1 969 : " Fluctuating Pressures in Spillway Stil lingBasins " , Ат. Soc. о/ Cii·il Engгs . . vol. 95, No. Н У.6 ( November).
23. Bozov1c А., 1 973 : " Yibration of the Bottom Outlet on the Bajina Basta dam ",l l th ICOLD, Q. 4 1 - R. 62, vol . 2 , Madrid.
24. B RUSH I N J., 1 984 : " Modelling Yibration and Vortex Formation at Piedro delAguila " , Water Рои·ег and Dат Constгuction ( January & Мау 1 984).
25. CAIN Р. and Wooo 1 . R., 1 98 1 : " Measurements of Self-Aerated Flow on аSpillway ", Ат. Soc. of Civil Engrs . . vol. 1 07, No. Н У. 1 1 ( November 1 98 1 ) .
26. С ЛM PBELL F. В. et а! . . 1 965 : " Boundary Layer Development and Spil lwayEnergy Losses " , Ат. Soc. of Civil Engrs . . Yol . 9 1 , No. Н У.3, Part 1 ( М ау) .
27. CARLIER М . et а/. (Comite Franyais des Grands Barrages), 1 979 : « Ouvragesd'evacuation de grande capacite », 1 3 " CIGB, Q. 50 - R. 6 1 , Yol. 3 , New- l)el hi .
28. С н ло Р. С . , 1 980 : " Tarbela Dam - ProЫems solved Ьу Novel Concretes " ,Civil Engineering ASCE ( December).
29. C нлvд R R I G., LOU I E D. S., COLEMAN Н. W., 1 979 : " Spill way and Tail raceOesign for Rais ing Guri Dam using Large Scale Hydraulic Model ". l 3thICO LD, Q. 50 - R. 1 2, Vol . 3, New-Delhi.
30. CoLA R., 1 966 : « Diffusione di un getto piano verticale in un bacino d'acquad'altezza li mitada » , L 'Eneгgia Elettrica. Vol . X L I I I ( Nov. 1966).
3 1 . COLGATE [) . , 1 972 : " Aeration Mitigates Cavitation in Spil lway Tunnels ", Ат. Soc. о/ Civil Engrs. National Water Research Meeting ( January).
32 . DAVIS С. У" SOR ENSEN, 1 979 : Handhook o/Applied Hydгaulics, Мс Graw Hi ll,3rd Edition, New- York.
33. l)EFAZIO F. G. and WEI С. У., 1 982 : " Simulation of Free Jet Trajectories Гог the Design of Aeration Devices on Hydraulic Structures ", 4th l nternational Conference on Finite Elements in Water Resources, Hannover, June 1 982 .
34 . DEL С А М РО А., TRI NCADO J . , 1 967 : " Some ProЫems in Operation 01· San Esteban Dam Spillways " , 9th I CO LD, Q. 33 - R. 33, Yol . 2, l stanbul.
35 . DESTENAY 1 . , B E RNARD J . , 1 968 : « Quelques exemples de degradation des bltons par cavitation dans des ouvrages hydroelectriques », La Houille Bla11cl1c. No. 213.
1 63
36. Dоuмл J . Н . , 1 97 2 : " Field Experience with Hydraulic Structures ", GeneralLecture I A H R Symposium on Flow l nduced Structural Vibration, Karlsruhe(August) .
3 7 . [)ROUIN р , Е . , BACAVE Р., 1 967 : « L e barrage d e Manicouagan 5 - Ouvrages de contrбle de la retenue » , 9е C IGB, Q. 33 - R. 32 , Yol . 2, I stanbul .
38 . D u FFAUT J . et а!. (Comite Fraщais des Grands Barrages), 1 97 3 : « Determination des crues de projet » , l l e C IGB, Q. 4 1 - R. 8 , Уо\ . 2 , Madrid.
39 . Есс н с R L. and S 1 EG ENTHALcR А. , 1 982 : " Spi l lway Aeratioп of the Sап RoqueProject ", Water Power and Dam Construction ( September).
40. FALVcY Н. Т., 1 982 : " Predicting Cavitatioп in Tunпel Spil lways " , Water Pmi·erand Dam Construction (August).
4 1 . FAU R E J., PUGN EТ L., 1 959 : « Etude de l 'a\imentation d'uп evacuateur en puits » , 6е Congres ital ien d 'Hydraulique et de construction hydraulique, Padoue, mai 1 959.
42. GALINDEZ А. , G u 1 N CA Р. , Lucлs Р. , AsPURU J . J . , 1 967 : '' Spil l ways iп а PeakFlow River ", 9th ICOLD, Q. 33 - R. 22, Vol . 2, I staпbul .
43 . GALPERIN R . S . , 1 97 1 : " Cavitation i n Elemeпts of Hydraul ic Structures апd M ethods o f C oпtrol l iпg it " , Hydrotechпical Coпstruction trans lated t'or ASC E, Gidrotekhnicheskoe Stroitel'stvo, No 8 (August).
44. GALPERIN R. S., OsкoLOv А. G. , 1 977 : " Protection against Cavi tatioп ЬуBoundary Flow Aeration " , puЫished in Russiaп in « Cavitation iп H ydraulicStructures » , Eneгgi_m. M oscow, U RSS.
45 . GALPERIN R. S . , OsкoLOv A.G. , ScM ENKOV У. М ., 1 977 : " Епегgу [)issi pationdowпstream the Gates ot' H igh H ead Water Outlets " , Proceedings 1 7th I A H RCongress, Yol . 5 , Baden-Baden, Germany.
46. GALPERIN R. S . , OsкoLOv А. G., SEMENKOV V. М . and TsшRov G. N . , 1 97 7 :" Cavitatioп of Hydraulic Structures " , Eneгgiya. M oscow, U RSS.
47. GARG J . М., TIAGI S. S. , 1 982 : " Eпergy Dissi pation for High ConcentratedFlow below Spi l lway in Narrow Gorges, Lakhwar Dam ", Transactions of thel nt . Symposium оп Layout of Dams in Narrow Gorges. Brazi l iaп Commiteeоп Large Dams, Vol . 1 , р. 22 1 , Rio-de-Janeiro.
48. G Pm L J., M ousтv J., 1 983 : (( Contribution а l ' etude de la protection coпtrel 'abrasioп et les chocs des ouvrages hydraul iques » , La Houille В!апс/1е. п6 3 /4.
49. G EORG E ROBERT L., 1 980 : " l mpiпgiпg Jets " , REC-ERC-80-8, US Bureau ofReclamation ( Dec.).
50. G н пт1 А., [)1 S 1 Lvo G., 1 967 : " l пvestigatioп оп the Runпiпg of Deep GatedOutlet Works from Reservoirs " , 9th I C O L O, Q. 33 - R. 48, Vol. 2 , [ stanbul .
5 1 . G I EZCNDANNER W . , H cNRY Р., 1 980 : (( Vidaпge d e foпd а graпde vitesse. Essais sur modele et protectioп du tuппel d'evacuation de Karakaya (Turquie) »,Ingenieuгs et Arcl1itectes Suisses ( 1 1 decembre).
52. H A K I M I А., С нлоu1 М . , M A RCHAND R., STERENBERG J . , 1 973 : (( Les ouvragesd'evacuation des crues du barrage de Moulay Youssef », 1 l e C I G B, Q. 4 1 -R. 74, Yo l . 2, Madrid.
53. H ALBRONN G . , 1 95 2 : (( Etude de la mise en regime des ecoulements sur lesouvrages а forte pente. Appl ication au proЫeme de l ' entrainement d'air » , La Houille Blancl1e (mai-juin-octobre-novembre 1 952 ) .
1 64
54. H A M I LTON W. S., " Preventing C<1vitation Damage to Hydraulic Structures "( in three parts ) , Water Poи:er and Dam Construction ( November 1 983, Oecember1 983 , January 1 984) .
55 . H ARTUNG F., 1 973 : " Gates in Spi l lways of Large Dams ", l l th ICO L D, Q. 41 -R. 72, Yol . 2, M adrid.
56. H A RTUNC; F. :шd HA USLER Е. , 1 973 : " Scours, Sti l l ing Basins and DownstreamProtection under Free Qyerfal l Jets at Dams ", l l th ICOLD, Q. 4 1 - R. 3 , Yo l . 2 .Madrid.
57 . H A RTU NC1 F. and KNAUSS J . , 1 976 : " Considerations f'or Spi l lways exposed to Oangerous C logging Conditions " , 1 2th ICO LO, Q. 47 - R. 2, Vol . 3 . Mexico.
58 . H A USLER Е. , 1 982 : « Spi l lways and Outlets with H igh Energy Concentration ", I ntern. Symposium on Layout of' Dams in Narrow Gorges, Rio-de-Janeiro. Tr:шsactions, Vol . 2.
59. H A USSI-:R R., FOUSECA F. et LA RIV I ERE R., 1 976 : ( ( Conception et exploitationdes ouvrages hydroelectriques en presence de glace », 1 2" C I G B, Q. 47 - R. 5. Yol. 3 , Mexico.
60. Н оuс нтоN О. L. , BoRCI-: О. Е . . 1 978 : " Cavitation Resistance of some SpecialConcrete ", А С/ Journal proc., Уо\. 75, No. 1 2 (December).
6 1 . HousтoN Kathleen L. , 1 982 : " Hydraulic M odel Study of Ute [)am Labyrinth Spi l \way ", G R-82-7 , US Bureau of Reclamation (August).
62. H U NC1 Тло S н EN ( Editor), 1 983 : " Frontiers in H ydraulic Engineering ",Hydraulic Specialty Conference on Cavitation Prevention and Mitigation, Ат.Soc. о/" Ci11• Епкrs . . 1 983 .
63 . IVEТIC Marks. У., 1 983 : " Pressure Fl uctuation in Sti l l ing Basins with af'terSki-Jump Spi l lways " ( Subject В-с), l ntern . Assoc. for Hydraulic Research.
64. JлвАRА М. А . , LECAS J . , 1 973 : " Selection of Spi l lways, Plunge Pools andSti l l ing Basins for Earth and Concrete Dams ", l l th ICO LD, Q. 4 1 - R. 1 7,Vol . 2, Madrid.
65 . JOH NSON R. В . , 1 967 : " Flood and M u ltiple Spi l lway Provisions at Large Earthen Dams ", 9th I C O L O, Q. 33 - R. 50, Yol . 2, M adrid .
66. JoH NSON R. В . , 1 973 : " Spi l lways Types in Australia and Factors aff'ecting theirChoice ", l l th ICO LD, Q. 41 - R. 46, Yol . 2, Madrid.
67. JOH NSON J . М., 1 977 : " Use of а Weakly Cohesive M aterial for Scale ModelScour Studies in Flood Spi l lway Design ", l 7th Congress IAH R. Yol. 4.
68. JOH NSON Н. А . and С н ло Р. С., 1 979 : " Rol lcrete Usage at Tarbela Oam ",Concrete International Design and Construction. Vol . !, No. 1 1 ( November).
69. J u Nsнov М" 1 982 : " The Layout and Energy !)issipation for Spi l lways ofseveral Hydroelectric Projects in Deep Val leys with Large Oischarge and High[)rops in H ead ", The Transaction of the Intern. Symposium on Layout of Oamin Narrow Gorges, Brazi lian Committee on Large Dams, Yol. 1 . р. 6 1 ,Rio-de-Janeiro.
70. KARAKI S. , Клz1 А . Н ., С нло Р. С. , 1 983 : " Performance of Air Slots in H ighYelocity Chutes at Tarbela Tunnel Outlet Works ", American Society of Civi lEngrs., H ydraul ic Division Conference, Cambridge, M assachusetts (August).
1 65
7 1 . KAVESHNI KOV А. т. апd LENTYAEV L. О., 1 978 : " Flow Aeratioп оп the Operatiпg Spil lway at the Sayaпo-Shusheпskoe Hydroelectric Statioп ", Gidrotekhnicheskoe Stmitel'stvo. Traпslated Ьу ASC E H ydrotechпical Coпstructioп.
72 . K ELLER R. J . , RлsтoG1 А. К., 1 977 : " Desigп Chart for Predictiпg Critical Poiпt оп Spillways " , Proc. o.f the Ат. Soc. o.f Civil Engrs., No. НУ. 1 2 ( December).
73. K EN N М. J. , 1 968 : " Factors iпПuепсiпg the Erosioп of Coпcrete Ьу Cavitatioп ", Tech пical поtе, 1 Ju ly 1 968, Coпstructioп lпdustry Research апdl пformatioп Associatioп ( C I RIA) , Lопdоп.
74. K ENN М. J . апd GAR ROD А. l). , 1 9 8 1 : " Cavitatioп Damage апd the TarbelaТuппеl Coll apse of 1 974 " , Proc. Insti. Civil Engrs. , Part 1 , 1 98 1 , 70 ( February),Proc. Insti. Civil Engrs" Part 1 , 1 98 1 , 70 ( N ovember) (Discussioп).
75. K I LPOLA I N EN J. Е., LOU NAMAA М. Е. , 1 976 : " Cl imatological Effects оп l)am[)esigп апd Coпstructioп оп the Kemi River ", 1 2th ICO LD, Q. 57 - R. 1 6,Yol . 3, Mexico.
76. KNAPP R. Т., l)ARLY J . W. апd Н л м м 1т F. G. , 1 970 : Cavitation. Мс Graw Hi l l ,New- York.
77. K NAUSS J . , 1 980 : « Talsperreп-Betriebseinrichtuпgeп ( Eпtlastuпgs-uпd Eпtпahmeaпlageп ) » , Oskar У. M il ler J пstitut, Oberпach/Walcheпsee ( Ameпagemeпt de l ' equipemeпt des barrages ) .
78 . K NAUSS J . , 1 985 : " Swirl i пg Flow ProЫems a t I пtakes ", Hpdraulic Structures Desixn Manual. M oпograph Н 1 , 1 пtеrп. Assoc. for Hydraulic Research .
79. LENCASTH А., 1 96 1 : " rree Overflow Spillways ", Eпgiпeeriпg апd DesigпPriпciple. Natioпal Laboratory Гог Civil Eпgiпeeriпg Studies, Report No . 1 74,Lisboп, Portugal.
80. L ENCASTRE А., 1 985 : « Etat des coппaissaпces sur le dimeпsioппemeпt desevacuateurs de crue des barrages » ( Аппехе : « Evacuateurs 3 jets croises » ),La Houille Blanche. N ° 1 ,
8 1 . LENCASTRE А., 1 985 : " C ross Jets - А Solutioп Гог 1 пcreasiпg Discharge Capacity or Spi l lways ", 1 5 th ICO LD, Q. 59-R. 26, Yo l . 4, Lausaппe.
82. LEV I N L., 1 965 : « Calcul hydraulique des coпduits d'aeratioп des vidaпges defond et dispositi f's deversaпts », La Houille Blanche. №' 2.
83. LEVIN L., 1 968 : Formulaire des conduites /orcees et conduits d 'aeration. Duпod,Paris.
84. LIEBL А., 1 973 : " High Pressure Slu ice Gates ", l l th I C O L I) , Q. 4 1 - R . 42,Vol. 2, Madrid .
85 . LюNпт1 F., C RAVIARI G. , 1 973 : « Prohlemes de coпstructioп et d'explo itatioп du barrage de Val Grosiпa еп rel atioп avec le coпtrбle des debits de crues et du charriage » , l l th I C O L D, Q. 4 1 - R. 78, Yol. 2, M adrid .
86. LowE J . 1 1 1 , BANGASH Н. О. апd Снло Р. С. , 1 979 : " Some E xperieпce withH igh Yelocity H ow at Tarbela l)am Project ", 1 3 th ICO LD, Q. 50- R. 1 3 , Yol. 3,New-l)elhi .
87 . Lowi:: J . 1 1 1 , С н ло Р. С. апd L U ECKER А. R., 1 979 : " Tarbela Service Spi l lwayPluпge Pool Developmeпt " , Water Power and Dam Construction ( November).
1 66
88. LOWE J . 1 1 1 and B I NG E R W. У., 1 982 : " Tarbela Dam Project, Pakistan ",Second Annual U SCO L D Lecture, Atlanta ( February).
89. LYSSENKO Р. Е . , BODEV У. D. , C н EPAJ K I N G . А., 1 97 1 : " Decrease in CavitationErosion lntensity for H igh Head Gates using the Supercaviting Structures ",1 4th Congress of I A H R, Vol . 5, Paris .
90. LYSSENKO Р. Е. , 1 983 : " Hydraulic, J)ynamic and Cavitation Characteri sticsof Curved Spil lways with the Flow along the Cei l ing " ( subject В-с) , XXthCongress of the l ntern. Assoc. for H ydraulic Research, M oscow.
9 1 . MA ITRE R., 1 95 2 : « Etude des conditions de fonctionnement des evacuateurs de surface aux ouvertures partiel l es des vannes », La Houille Blanche, № 2 .
92 . Млмп J . , LLOPIS N., D'ANG E:LO А. , 1 983 : « Les revetements anti -usure desbarrages du Rhone » , La Houille Blanche, № 3/ 4.
93. M ARТI NS R., 1 973 : " Contribution to the Knowledge on the Scour Action of'Free J et on Rocky River Beds ", l l th I C O L D, Q. 4 1 -R. 44, Vol. 2, M adrid.
94. M ARТI NS R., 1 97 5 : " Scouring of' Rocky Riverbeds Ьу Free-Jet Spi l lways ",Water Power and Dam Construction ( Apri l ) .
95 . M ASON Р. J . , 1 983 : " Energy Dissipating C rest Splitters for Concrete Dams ",Water Power and Dam Construction ( N ovember).
96. MлsoN Р. J., 1 984 : " Erosion of Plunge Pools Downstream of Dams due tothe Action of Free-Trajectory Jets ", Procedures и/ the lnstitution о/ Ci�·i/ Engineers. Part 1 (Мау).
97. M AТIAS Rлмоs С . , 1 982 : " Energy Dissipation of Free-Jet Spil lways. Basis forits Study on H ydraulic Model ", The Transaction of the Int . Symposium onLayout of Dams in Narrow Gorges, Brazilian Committee on Large Dams,Vol. 1 , р. 263, Rio-de-Janeiro.
98. M E I DA L Р., WABSTER J. L. , 1 973 : " Mica : One of the World's LargestStructures ", Water Power and Dam Construction. Part. 2 ( July).
99. M EI DAL Р., WABSTER J. L., 1 973 : " Discharge Faci lities for M ica Dam " , l l thICO LD, R. 50-Q. 4 1 , Vol . 2, Madrid .
1 00. M YERS У. А . , 1 967 : " Meteorological Estimation o f' Extreme Precipitation of' Spi l lway Design Flood ", US Weather Bureau, Techn. M emo Hydro. 5 .
1 О 1 . N о v лк Р., 1 984 : Developments in Hydraulic Engineering, Elsevier Applied Science PuЬ\ ishers Ltd, Vol . 2 , England.
1 02 . OвERLE R., Du влs С . et а! . . 1 967 : « Protection contre l'ensaЫement du bassin d'accumulation de l 'amenagement de la Massa » , 9е C I G B, Q. 33 -R . 37, Yol. 2 , I stanbul .
1 03 . O LIVIER Н . , 1 966 : " Flood Control Technique f'or Reservoirs with Comhiпed Free and Gated Spil lways ", 6th Congress I nt . Comm. o f' I rrigation and Drainage, Q. 22, New- Delhi .
1 04. OsкoLкov А . G., SEMENKOV У. М., 1 979 : " Experience in developing M ethods f'or Preventing Cavitation in Excess Flow Release Structures " , Gidroteklmicheskoe Stroitel ' Stvo. Hydrotechnical Construction N° 8 translated in English Ьу American, Soc. of Civil Engrs . , р. 754-76 1 , 1 980.
1 6 7
1 05 . OsкoLкov А. G., SEM ENKOV У. М ., 1 979 : " Experience in Designing and Maintenance of Spillway Structures on Large Rivers in the USSR " , l 3th I CO L [), Q. 50-R. 46, Yol . 3 , N ew- Delhi.
1 06 . Рлт and R E I N H ARDT, 1 979 : " Erosion of Concrete " , Heron, No 3 , N etherlands.
1 07 . РЕТЕRКА А. J . , 1 953 : " The E ffect of Entrained Air on Cavitation Pitting " , Joint Meeting Paper А 1 RH-ASCE. Minneapolis, Minnesota I nternational Hydraulics Convention (August).
1 08 . РЕПRКА А. J., 1 97 8 : " Hydraulic [)esign of Still ing Basins and Energy [)issipators ", US Bureau of Reclamation, Engineering Monograp/1 No 25, Revised edition.
1 09 . PI NTO Nelson L . de S. , N Ы DERT S . Н ., Отл J. J. , 1 982 : " Prototype and Laboratory Experiments on Aeration at H igh Yelocity Flows " , Water Power and Dam Construction, Part 1 ( February), Part 2 ( March).
1 1 О. PI NTO Nelson L. de S. and №::I DERT S. Н . , 1 982 : " Model Prototype Conformity in Aerated Spil lway Flow ", I ntern. Conf. on the Hydraul ics M odeling of CiYil Engineering Structure B H RA Fluid Eng., England (September) .
1 1 1 . PI NTO Nelson L. de S. and N E I DERT S . Н . , 1 983 : " Evaluating Entrained Air Flow through Aerators " , Water Power and Dam Construction ( August).
1 1 2. PI NTO Nelson L. de S. and N EI DERT S . Н . , 1 983 : " Air Entrainment through Aerators Mechanism and Air Flow Evaluation ", lntern. Assoc. for Hydraulic Research, Seminar 3 , M oscow.
1 1 3 . PI NTO Nelson S . de L., 1 984 : " Model Evaluation of Aerators in Shooting Flow ", Symposium on Scale E ffects in M odelling Hydraulic Structures. Esslinger am N eckar, Germany (September 3 rd-6th).
1 1 4. Роsт G . et а/. (Comite Fraщ:ais des Grands Barrages), 1 979 : « Quelques proЫemes particuliers poses par les deYersoirs а grande capacite : tapis de protection, dissipation d'energie par deПecteurs et aeration, cavitation produite par les ecoulements ;;i grande vitesse » , 1 3" C I G B, Q. 50- R. 38, Vol . 3, New-[)elhi .
1 1 5 . PRATT Н . К . , 1 967 : " The Portage Mountain Project Spi l lway and Low Levcl Outlet Works ", 9th ICO L D, Q. 33-R . 29, Vol. 2, l stanbul.
1 1 6. QU I NТELA А. С., FERNANDES J . de S . et а/. , 1 979 : « Barrage de Cabora Bassa -ProЫemes poses par le passage des crues pendant et apres la construction » ,1 3е C I G B, Q. 50-R. 4 1 , Vol. 3 , N ew-Delhi .
1 1 7 . QU I NTELA А . С., J ELLALI М. et а/. , 1 979 : « L'evacuateur de crues et les vidanges de fond du barrage de M 'Jara » , 1 3 " C I G B, Q. 50- R. 40, Vol . 3, New-J)elhi .
1 1 8 . QU I NПLA А . С . , 1 980 : " Flow Aeration to Prevent Cavitation Erosion " , Water Power and Dam Construction (January).
1 1 9 . QU I NTELA А. С . , C RUZ А. А., 1 982 : " Cibora Bassa Spi l lway Conception, Hydraulic Model Studies and Prototype Behaviour ", The Transaction of' the l nt. Symposium on Layout of Dams in Narrow Gorges. Brazilian Committee on Large Dams, Yol . 1 , р. 30 1 .
1 20. REGAN R. Р., M u Ncн А. У., SCHRADER Е . К . , 1 979 : " Cavitation and Erosion Damage of S luices and Sti l l ing Basins at two High H ead Dams ", 1 3th J C O L D, Q. 50- R. 1 1 , Yol . 3, New- Delhi .
1 68
1 2 1 . RONE Т. J " 1 977 : " BafПed Apron as Spil lway E nergy Dissipator ", Proc. of the Ат. Soc. of Civil Engrs" Vol. 1 03 , No. НУ. 1 2 (December).
1 22 . RIQUOIS М" PFAFF, TERM ISSIAN et а/" 1 967 : « ProЬ\emes poses par l 'exploitation et l 'eлtretien des organes d'evacuation des barrages. Enseignements tires » , 9е C IG B, Q. 33-R. 28, Vol . 2, I stanbul .
1 23 . RoBERTS D. F. , 1 977 : " Eпergy Dissipation Ьу Dam Crest Spl itters " , Trans. Sout/1 A/rican lnst. о( Civil Engr.1. ( November).
1 24. ROBI:RTS С . Р. R. , 1 980 : " Hydraulic Desigп of Dams " , Department of Water Affairs. Division of Special Tasks, RepuЬ\ic of South Africa (July).
1 25 . RosANOV N . Р., M oYs Р. Р., Рлsн коv N. N., 1 967 : " Research of Yacuum апd Cavitatioп Characteristics of Elemeпts of Hydrotechпical Structures " , 1 1 th Congress I A H R, R. I . 33, Leningrad.
1 26. RU NDG R EN L., 1 973 : " Desigп Floods апd Accepted Risks of Fai lure ", 1 1 th I C O LD, Q. 4 1 - R . 1 1 , Vol . 2 , Madrid.
1 27 . Russa S . О. апd BALL J. W., 1 967 : " Suddeп- Eпlargemeпt Eпergy Dissipator for Mica Dam " . Рте. о/ t/1e Ат. Soc. о( Ci1•i/ Engrs" No. Н У.4.
1 28 . R ussE L S. О" SHEENAN G . J. , 1 974 : " Effect of Entrained Air on Cavitation Damage " , Сапаdiап Journal of Civil Eпgrs . , Vol . 1 .
1 29 . SANC H EZ B R I B I I:SCA J . L. , CAPELLA YISCAINO А., 1 973 : " Turbuleпce Effects оп the Liпiпg of Sti l l iпg Basins " , l l th ICO LD, Q. 4 1 - R . 83, Yol. 2 , Madrid .
1 30. SANC H EZ В R 1 в 1 1:sсл J . L. , 1 979 : " Behaviour of Spi l lways iп Mexican Dams ", Universidad N aciona\ Aut6noma de M exico. Contribution to the 1 3th I COLD, New-Delhi. I пstituto de I ngenieria (October), Е-40.
1 3 1 . SANCHEZ B R I B I ESCA J . L. , FU ENTES M ARI LES О. А., 1 979 : " Experimeпtal Analysis of M acroturbulence Effects on the Lining of Stilling Basins " , 1 3th I CO LD, Q. 50- R. 6, Yol . 3, New- Delhi .
1 32 . SдNCH EZ B R I B I I:SCA J . L. , 1 979 : " Behaviour of Spil lways iп Mexican Dams ", 1 3th l C O L O, Q. 50, Discussion No. 1 4, р . 528, Vol . 5, New-Delhi .
1 33 . SCH U LTH ESS D., 1 985, " Yortex Tower for El Cajon Spi l lway I пtake " , Water Power and Dam Constr. (Ju ly).
1 34. SEM ENKOV У. М . апd LENTYAEV L. D., 1 973 : " Spil lway with Nappe Aeratioп " , Gidrotekhnicheskoe Stroite/ ' Stvo, H ydrotechnical Coпstructioп ( М ау).
1 3 5 . SI:M EN KOV У. М ., LENTYAI:V L. D., 1 973 : " Spi l lway Dams with Aeration of the Flow over Spil lways ", l l th I C O LD, Q. 4 1 , Discussioп, а separate рарег, Vol . 5, Madrid .
1 36. S E M E N KOV У. М ., 1 979 : « Yidaпges et evacuateurs de crues de grande capacite » - Rapport General, 1 3е C I G B, RG Q. 50, Yol . 5, New-Delhi .
1 37 . SндR\1А Н. R., 1 976 : " Air- Eпtraiпmeпt iп High H ead Gated Conduits " , Proc. о/ t/1e Ат. Soc. о( Civil Engrs" Yol . 1 02, No . Н У. 1 1 .
1 38 . Sмпн Н . А., 1 973 : " А Detrimeпtal Effect of Dams оп Eпviroпmeпt : Nitrogen Supersaturatioп ", 1 1 th I C O L O, Q. 40 - R. 1 7 , Vol . 1 , Madrid.
1 39 . SNYDI:R F. F., 1 964 : " Hydrology оГ Spil lway Design Flood = Large Structures, Adequate [)ata " , Proc. oftl1e Am. Soc. of' Civil Engrs" Yol. 9 1 , No. Н У.3 .
1 40. SoкoLOv А. А. , RANTZ S . Е . апd Rос н Е М., 1 976 : " Flood F low Computatioп. Methods compi led from World Ехрегiепсе ", The U N ESCO Press, Pari s .
1 69
1 4 1 . SPURR К. J . W., 1 985 : '" Eпergy Approach to Estimatiпg Scour dowпstream of а Large Dam " , Water Power and Dат Construc. (July).
1 42 . STRASSBURGER А. G., 1 973 : " Spillway Eпergy Dissipator ProЫems ", l l th ICOLD, Q. 4 1 - R. 1 6, Yol . 2, Madrid.
1 43 . STRAUB L. G., AN DERSON А. G., 1 960 : " Sel f'-aerated Flow in Ореп Chaппels " , Transactions о( tl1e Ат. Soc. о( Civil Engrs . . paper 3029, Yol . 1 25 , Part 1 .
1 44. Suzuк1 У., S д K U RA I А., 1 973 : '" А Desigп o f' а Chute Spi l lway Joiпtly Servi пg as а Roof Slab of а Hydropower Statioп апd its Review оп Vibratioп duriпg Flood ", l l th ICO LD, Q. 41 - R. 2 1 , Yo l . 2, Madrid.
1 45 . TORAN J . , 1 967 : " The Dowпstream Yiew-poiпt оп Spi l lway Safety апd Ecoпomics " , 9th ICO LD, Q. 33 - R. 52, Yol. 2 , l staпbul .
1 46. ToRAN J . , 1 967 : « Dispositioпs temporaires et permaпeпtes pour coпtrбler les apports et le пiveau de la reteпue des barrages » - Rapport Geпeral, 9е C I G B, Q. 33, Yol. 6, р. 225-336, I staпbul .
1 47 . TsEDROV G . N. апd NдZAROVA R. L. , 1 977 : " Protectioп agaiпst Cavitatioп iп Тuппеl Spillways Ьу Bouпdary Layer Aeratioп " , Gidmteklmicl1eskoe Stгoitel " Stvo, No. 8, Hydrotechпica\ Coпstructioп.
1 48 . U S BU REAU OF REC LAMAТION, 1 974 : Design о/ Sта// Daтs. ReYised Repri пt, Washiпgtoп .
1 49. U N IТED N ATI ONS, 1 969 : " Assessmeпt o f' the Magпitude апd Frequeпcy of Flood Flows " , Water Resources, Series No. 30, New-York.
1 50. YASI L I U М . F., 1 979 : ProЬlemes particuliers de quelques evacuateurs de graпde capacite, 1 3е C I G B, Q. 50 - R. 28, Уо\ . 3, New-Delhi .
1 5 1 . VдSS I L I EV О . F. , B u K RI::YEV У. 1 . , 1 967 : '" Statistical Characteristics of Pressure Fluctuatioп iп the Regioп of H ydraulic Jump ", 1 2th I A H R Coпgress, Yol . 2. Fort-Colliпs, Colorado.
1 52 . YERCON М . , 1 973 : « Coпtrбle des debits et de la dissipatioп d'eпergie репdапt la coпstructioп et apres la mise еп service » - Rapport Geпeral, 1 1 е С 1 G В,
Q. 4 1 - RG, Yol. 4, Madrid.
1 53 . YISHER О., 1 982 : « Hochwassereпtlastuпgeп brasi liaпischer Talsperreп : drei Eiпdrucke vom ICO L J)- Koпgress 1 982 iп Rio-de-Jaпeiro » (The Large Brazilian Spillways), Wasser, Energie, Luft, n° 1 О.
1 54. VoтRUBA L., 1 976 : « L'iпflueпce du changemeпt anпuel de temperature sur les barrages et les reservoirs » , 1 2е C I G B, Q. 47 - R . 20, Yol. 3 , Mexico.
1 55 . W дG NER W. Е. , 1 967 : " Gleп Саnуоп Dam Diversioп Тuппеl Outlets " , Proc. о/ the А т. Soc. of Civil Engrs., Vol. 93, No . НУ.6 (November).
1 56. W дGNER W. Е. апd JдBARA М . А., 1 97 1 : " Cavitatioп Damage dowпstream from Outlets Works Gates " , 1 4th I A H R Coпgress, R. 2 . 1 4, Yol. 5, Pari s .
1 57 . WISNER Р., 1 965 : « Sur le rбle du critere de Froude daпs l 'etude de l 'eпtraiпemeпt de l'air par les couraпts а graпde vitesse » , 1 1 е Coпgres А ! R H , R. 1 . 1 5 , Leпi пgrad.
1 58 . WISNER Р., 1 967 : " Hydraulic Desigп for Flood Coпtrol Ьу High H ead Gated Outlets " , 9th ICO LD, С 1 2, Yol . 5, I staпbul .
1 59. Wooo 1 . R. , AcкERS Р . апd LoVELESS J " 1 983 : ' " Geпeral Method for Critical Poiпt оп Spil lways " , Techпical Note, Proc. о/Ат. Soc. ofCivil En[;rs" Yol. 1 09, No. Н У.2.
1 70
1 60. Wooo 1 . R., 1 983 : " Uniform Region of Self-Aerated Flow ", ASCE - Journal of Hydraulic Engineering, Vol. 1 09, No . 3 ( March).
1 6 1 . Wooo 1. R., 1 984 : " Air Entrainment in High Speed Flows " , Symposium on Scale Effects in Modelling H ydraulic Structures - Esslinger am Neckar, Germany (September 3-6).
1 62 . WoRLD WATER, 1 979 : " Cavitation Casts Doubt on Karun Spillway Design ", World Water (January).
1 63 . XEREZ А. С . , GRANGER Р1 Nто Н ., 1 967 : « ProЬ\emes hydraul iques dans les barrages voйtes et voйtes m ultiples » , 9е C IGB, Q. 3 3 - R. 1 8, Vol. 2, I stanbul.
1 64. Youo1sк1Y G . А., LIAKHTER У. М., 1 968 : " H ydrodynamic Loads on Protection Elements downstream of Overspill Dams ", Energiya. M oscow.
1 65 . ZAG USТI N К . and CASTI LLEJO RoN N ., 1 983 : " Model Prototype Correlation for Flow Aeration in the Guri Dam Spil lway " , l ntern. Assoc. for H ydraulic Research , 20th I A H R Congress, Moscow, 1 983 .
1 7 1
ANNEXE 21 APPENDIX 2
COUNTRIES WHICH HAVE ANSWERED ТНЕ ENQUIRY ON SPILLW А У&
Algeria Australia Austria Belgium Brazil Canada China Czechoslovakia France Germany (Fed. Rep. of) Indonesia Japan Mexico South Africa Spain Switzerland Thailand USA Venezuela Zimbabwe
Imprimerie de Montligeon 6 1 400 La Chapelle Montligeon
Depбt legal : fevrier 1 987 № 1 3 149
ISSN 0534-8293 Couverture : ТIL Т
1 72
Copyright © ICOLD - CIGB
Archives informatisées en ligne Computerized Archives on line
The General Secretary / Le Secrétaire Général : André Bergeret - 2004
_______________________________________________________________
International Commission on Large Dams Commission Internationale des Grands Barrages 151 Bd Haussmann -PARIS –75008
http://www. icold-cigb.net ; http://www. icold-cigb.org
