AT
THE
SOVIET
DURABILITY
OF
NATIONAL
DAMS
AND
COMMITEE
ESTIMATION
ON
OF
LARGE
THEIR
DAMS
BEHAVIOR
V. V. Stol'nikov
UDC 627.821.2.004:006.3(100)
The problem of estimating the behavior of dams of various types after a certain interval of their operation attracts continously the attention of designers and constructors all over the world. At the Ninth International Congress on Large Dams*, 53 papers from 26 countries were presented on this problem (No. 34), which was the subject of a detailed discussion. The papers examined concrete as well as earthfill and rockfill darns. The most compreheusive was the analysis of problems of concrete deterioration. From the USSR a paper was presented by S. Ya. Eidelman (VNIIG) devoted to the results of instrumental observations at the Bratsk Hydroelectric plant,"The 50th Anniversary of the Great October." The problem was examined at the Congress under various aspects, most important of which were the design of dams, the observation of dams during the period of operation, repair and reconstruction, and the general problems of organizing and systematically controlling the behavior of dams of various types. As noted in a series of papers and emphasized by the general reporter D. Bullet (U.S.A.), the causes of deterioration and premature wear of dams are generated by insufficient investigation and poor study of the construction region, and also by design deficiencies, and particuiarly, by the incorrect selection of the dam type, and by the l a c k of adjustment of the dam to the actual conditions in which the construction works. Thus, according to the Czechoslovak National Committee, during the last 15 years several failures involving dams of heights between 5 and 21 m occurred because of the poor study and insufficient bearing capacity of the foundations. A failure involving a flood protection dike of 5.1 m height on the Danube occurred for the same reason. The comprehensive consideration of the c l i m a t i c and operation factors which produce deterioration of the concrete in the operation process is very important. The modern state of the science of concrete and the level of knowledge of its properties make it possible to ensure a long and safe life service of hydraulic structures. However, 11 papers reported the destruction of concrete by frost because of interaction between the cement alkali and the aggregate, incorrect selection and specification of concrete requirements, and poor quality of concrete preparation and compaction during execution of the work. The papers emphasized that these destructive processes can be reduced to a minimum by the correct selection of the cement, aggregate, and additives, the ratio between the component parts of the concrete mix, careful casting, compaction, and curing of the concrete. The papers stressed especially the requirements of using aggregates of high quality and air entraining agents, and also of thermal protection of the concrete in the earlier stages to eliminate possible damage during the construction period. Interesting examples of constructive protection of thick wall concrete dams against destruction by frost were given in the paper No. 28 (Norway). The severe c l i m a t i c conditions of that country, especially in the central regions, are characterized by frequent variations from thawing to freezing conditions at very low t e m p e r a t u r e s ( - 3 0 e C ) and create a difficult operation regime for concrete in dams. Investigations on 132 existing dams were carried out in Norway, and these enable drawing a few conclusions. The types of dams used most often in Norway are the A m bursen (Fig. 1) and arch (Fig. 2) dams. Characteristic for both types of dams is the presence of a thin insulating wall intended to eliminate the sudden t e m perature changes at the downstream part of the dam and reduce the danger of damage by frost. Observations showed that for cases when the surrounding temperature fell down to -80~ the air temperature between the dam and the insulating wall only fell under -10"C infrequently. In a few cases heating systems were installed in the space b e tween the insulting wall and the dam. Very positive results were obtained in Norway by using air entraining agents, which are considered in Norway the most reliable method of increasing the durability of dam concrete subject to frost conditions. In repairing the surface of concrete dams damaged by frost, good results have been obtained by using a mixture of epoxy tar and fine quartz sand. * See the author's paper in the journal Gidrotekhnlcheskoe Stroitel'stvo, No, 7, 1968. Translated from Gidrotekhnicheskoe Stroitel'stvo, No. 1, pp. 44-48, January, 1969.
68
DURABILITY OF DAMS AND ESTIMATION OF THEIR BEHAVIOR
69
!
a
a
5.0
1~ '/ --llr 1 /e
b
C
Fig. 1. Norwegian Ambttrsen type dam. a) Vertical section; b) horizontal section; c) supports system diagram. 1) Slab of various thickness T = (0.8 + 0.02 h) m; 2) buttresses; 3) insulating wall; 4) buttress bracing struts; 5) inspection gallery; 6) joint.
/ Fig. 2. Norwegian arch dam. a) View from downstream; b) plan; c) cross section T = (0.8 + 0.04 h) m. 1) Overflow spillway; 2) insulating wall; 8) joints. In connection with earth dams it was noted that the incorrect division of the dam body in zones may result in operation diffcuities, and sometimes in failure. When examining the problems of infiltration through earth dams it is necessary to take into account both chemical action of water (as a solvent) and the hydrodynamic action resttlting from the transportation of solid particles. The construction methods also are very important. Hydraulic fill dams are considered in paper No. 25 (USA) as unsafe, especially in seismic regions.
The relationship between the safety and durability of the dams and the carefulness of the investigations and correct evaluation of their results was also emphasized in the recommendations of the Congress on the problem u n der consideration. Another aspect of problem No. 34 dealt with observations on the behavior of dams from the moment of their completion and start of operation; this may include visual observations, observations by means of built-in and other instrumentation, and by means of supersonic instrumentation. Attention was called to the value of visual observations when these are carried out by experienced specialists. The apparition of cracks, rapid settlements, or the suddent start or increase of the i n f l t r a t i o n water flow may signal the presence of vast dam damage. In all such cases of damage it is necessary to investigate and establish their cause rapidly and carefully. It was noted at the Congress that in many cases observations were started immediately after completion of construction, but after a certain time the interest in pursuing them diminished, the responsibility for the inspection of the dams was transferred to people lacking interest in the problem, and many signs of defective dam conditions remained unobserved or were observed quite late. Thus in paper No. 14 (South Africa) it was noted that the 40 m high Churchili Dam, a thin multiple arch dam consUueted during 1940-1948 on the Kromme River, developed in 1959 a network of cracks on its surface which was observed by visual inspection. The cause of this damage was incorrect selection of the concrete aggregate, as demonstrated later experimentally. It was noted that the instrumentation type, its accuracy and operation re!lability present a large variation. Instrumental observations must start at the earliest stages of construction and continue after completion. The results of the observation must be examined by independent experts.
~/0
V. V. STOL'NIKOV
600
I:
0
/
.5o0 fl 0
c~
9
~o0
.]
9
0
%9
9
200
o
Examples were given of cases where instrumental observations have prevented serious damage. There was also a case where false evaluation of the observational data led to a failure.
-/o ,Q
sao
The observational data obtained must be compared with the project computations to complete the knowledge on the forces acting on the dam under actual operating conditions and improve the computation methods.
"'} .)./..
In the practice of Electricit6 de France (Paper No. 30) the results of i n strumentational observations are processed by computers.
.
9
9
i oo
/ ~ lO0 2000 3000 ~OOO 50@0 0 o Propogation velocity of ultrasounds, m / s Fig. 3. Relationship between compressive concrete strength, obtained from measurements of propagation velocity of ultrasounds and from the results of testing of specimens drilled from the dam body.
Paper No. 35 (Switzerland) describes the results of observations on the Mauvoisin dam. From the time of its completion in 1957 the reservoir was filled nine times. The dam deformations were measured by plummets, clinometers and geodetic instruments. In the earlier years of operation computation results were in good agreement with the observed data. However, plastic deformations of the rock and settlements of the dam were observed later and these continued up to the time of the writing of the paper. Some of the papers presented examples of the utilization of supersonic instrumentation for the evaluation of dam conditions. Paper No. 5 (USA) describes the 20-year experience in the utilization of supersonic methods, by which 29 dams were investigated. The general consensus formulated as a result of discussions was that the modern supersonic methods are a useful supplement to other, more reliable methods of determining the quality or behavior of dam concrete. The supersonic method can be used in the first place for establishing the portions of the construction where the concrete is of poor quality. It was noted at the Congress that these methods are not yet sufficiently reliable for determining the strength characteristics of the concrete.
Extensive studies of the application of supersonic methods were carried out in Sweden (Paper No. 22). The significant scatter of the results, as observed in the paper, is the basis for the conclusion that the estimation of concrete strength by means of supersonic methods is not sufficiently reliable (Fig. 3). Paper No. 8 (France) gives the results of the supersonic investigation of the vault of the Roseland Arch Dam (upper Savoie) in two horizontal sections, and notes that the concrete was found to be very homogeneous according to the indications obtained regarding the variation of the modulus of elasticity. Paper No. 49 (Italy) demonstrates that utilization of the methods and means of geophysics and especially the geoseimic method are fully applicable to the systematic investigation and general evaluation of concrete structures. Most mccessful appeared to be the method based on the variation elastic wave propagation and measurement of the free vibrations of the construction vault. Investigations on dams were carried out also by means of a special system of geophones. This system makes it possible in particular to record the variation of the sonic waves in the concrete structures as a function of the water level in the reservoir. Apparently the geophones can be used also for estimating the concrete porosity and its variation with time. A series of papers was devoted to the description of the repair of dams of various types. The general reporter paid much attention to the description of a method to eliminate infiltration through the structure foundations based on the construction of a trench along the front of the structure. This method was used successfully at several large dams in a number of countries. Although most often its utilization was stipulated in the dam projects, it was used also at existing structures. The method consists in the creation of trenches with vertical walls completely supported by hydrated bentonite paste. When the trench has reached the desimed depth it is back filled and the fill material becomes impervious because it is covered by bentonite particles. The above-described method for the elimination of infiltration through the foundation is not too dissimilar to the ICOS-Veder method utilized in Italy and described in the paper No. 34 (Finland). This method was applied for the construction of the cutoff wall in the foundation of the earthfill darns of the Montta hydroelectric plant on the Oulu River and Seitakorva hydsoelectrie plant on the Kemi River. Wells having a diameter of 50-60 cm were drilled at 110 and 120 cm intervals and filled with concrete, following which interstitial overlapping wells were drilled and fiUed with concrete.
71
DURABILITY OF DAMS AND ESTIMATION OF THEIR BEHAVIOR
" T -
:~:::. a
9 &.
~..~ ::";'!] 9
.'v 7"~j- ,:
.- ..;."
Fig. 4. Sealing of the Gurledan Dam (France) joints: a) Before recomtruction; b) after reconstruction. 1) Reinforced concrete sealing plug; 2) sealing cement piaster; 3) bituminous putty; 4) "inertol" paint; 5) drainage shaft; 6) basic sealing copper plate; 7) auxiliary copper strips; 8) anchor bolts.
.'.~"~
Fig. 5. The facing of the Maentwrog Dam (England): 1) Original concrete of the internal zone; 2) new facing; 3) drainage system; 4)bitumen fiberglass insulation; 5) of foundation grouting.
The repair of external darn surfaces also received attention in the papers presented at the Congress. In all the surface was cleaned by the usual method using sand and scrapers. In some cases the surface was then made more nearly water tight with gunite. Metallic linings, reinforced concrete, asphalt, and epoxy were also used. Paper No. 19 (France) described the 20-year experience in repair work carried out by Electricit~ de France on 280 dams. There are data on the good, long service life of m e t a l l i c linings. The paper describes reconstruction of the 54-m high G = l e d a n darn, built in 1980. Since leakage was observed in the vertical joints, these were reconstructed with positive results (Fig. 4). Interesting examples of repairs using e p ~ y coatings are given in Papers No. 19 (France) and 21 (Sweden). The latter describes reconstruction of the 22-m high Suorva multiple arch darn on the Lule River, north of the attic c i r cle. During the winter this darn is emptied and the insufficient vertical reinforcement in the lower part of the dam resulted in a very large number of horizontal cracks through which the water leaked. The reconstruction work was c a = i e d out in 1962 and seven different types of plastics received from England and the USA were used for sealing. The American modifier "Thioko1" was introduced into the plastic composition. About 3.5 k g / m z of epoxy were used, forming a 3 - r a m thick layer, and when a mixture of epoxy and neoprene was used, the consumption was about 2.5 k g / m z of epoxy and 2.5 k g / m z neoprene. In the English experience, repair of the 27-m high Maentwrog arch darn in North Wales, with a length at the crest of 1000 m, constructed in 1020, deserves attention (Paper No. 13). The dam concrete had a 6:1 mix ratio by volumes in the internal zone and 4:1 in the external zone having a depth of 50 cm. Unwashed, crushed sand was
72
V . V . STOL'NIKOV
used and also "plums". The concrete mix has a dry, rigid consistency. The casting was carried out without vibration according to the practice of those days. In 1928 the reservoir was filled and soon through the construction and horizontal joints drops of leached lime appeared. From 1938 to 1941 grouting of the d a m was carried out, but without success. In 1958, capital works were carried out (Fig. 5) following which the infiltration was significantly reduced, but it still takes place, presumably along the metallic bars reinforcing the bitumen fiberglass insulation. The d a m is in operation at the present time. As was noted at the Congress, m u c h attention was given in recent years to the systematic and continuous inspection of the dams. This was prompted by the failure of the Vajont D a m in Italy and Baldwin Hill D a m in the USA. In particular, the U. S. Federal Power Commission has issued a regulation requiring the periodical inspection (each five years) of the safety and behavior of dams, carried out by independent experts. In Czechoslovakia there are two organizations which insure inspection of large dams and hydraulic structures. In Yugoslavia the government issued regulations for the technical inspection of dams in 1965. The essential elements of inspection are establishment of the required instrumentation during the construction period, the carrying out of continuous observations, and the analysis of results of these observations by qualified specialists. In the course of the discussion on Problem No. 34, apart from the representatives of various countries, five earlier elected official experts also took the floor: L. Loussinr (France) G. Press (FGR), R. Urbistonde (Spain), S. S h u m (Switzerland) and V. V. Stol'nikow(USSR). In the expertise closure by V. V. Stol'nikov (USSR) an analysis of the papers was m a d e and he also raised the problem of the design service life of dams of various types under various climatic conditions. It was emphasized that in order to ensure the durability of the structures it is necessary to observe the principle of equal resistance capacity of all parts against negative factors. The list of the general requirements imposed on concrete must include beside those regarding its strength, specifications on its water absorption, settling, swelling, deformation c a p a city, and other properties which must be considered in the selection of the concrete mix.
Presenting later a paper from the International Committee on the Concrete of Large Dams, Professor V. V. Stol'nikov made a series of recommendations for the evaluation of concrete mixes for reinforced concrete dams and the method of evaluating the properties of concrete in the structures. The considerations and propositions set forth in the expertise closure received general approval and were i n cluded in the conclusions and recommendations of the Congress on the results of the examination of Problem No. 34, some of which are given in the following: CONCLUSIONS 1. The observations and measurements on infiltration, dam displacement, and damage of the materials from which the d a m was built must be carried out condnuonsly, and in any case, at intervals not longer than 5-7 years. 2. The built-in instrumentation must be installed during construction, the readings must start with the start of the construction and continue as long as construction is carried on. 3. The concrete for the d a m and related structures must be of the best possible quality and correspond to the severity of the local actions on it, since the capacity of the concrete to resist destruction is the most important guarantee of the safety of the dam. The body of the dam must be divided into zones and each zone must be cast from concrete having the amount of cement and the composition which is required by the intensity of the destructive factors acting upon it.
4. The problem of determining the concrete strength and its other properties in existing dams is most k n portant and this has been insufficiently investigated up to now. 5. The most r e l i a b l e method consists in extracting specimens of the required size from the dam body and testing them. The specimen extraction can be carried out by linear drilling or by taking cylindrical cores. The smallest dimension of the specimen must be at least twice the largest dimension of the aggregate particles in the concrete.
6. At the present time testing of the specimens taken from the dam body is carried out basically to determine the compressive strength, unit weight, and the permeability.
DURABILITY OF DAMS AND ESTIMATION OF THEIR BEHAVIOR
73
Taking into account the complex character of the action of external factors on concrete structures, the testing must include a larger sphere of concrete properties (the elasticity modulus, Poisson's ratio, tensile strength, deformability, leaching capacity, water absorption, resistance to freezing, etc). Analyses of the concrete should also be carried out. 7. The approximate determination of the compressive strength of the concrete in the structure can be carried out with sufficient accuracy by means of transfer coefficients from the test data obtained from the control specimens to those corresponding to the dimensions of the structure, provided that the concrete in the dam was cast under good quality control, with a m i n i m u m of stratification, and carefully compacted. Transfer coefficients can be established not only with respect to strength, but also to the magnitude of the permeability and frost resistance. 8. The utilization of supersonic measurements is most adequate for establishing the portions of poor concrete and especially the cracks in concrete structures. Estimation of concrete strength by this method is still unreliable. 9. The utilization of percussion methods is adequate for estimating the surficial strength of the concrete and must be considered as an auxiliary means for the visual inspection of the structures. 10. The methods of estimating the strength and other properties of the concrete in existing dams needs further improvement and the development of new instruments for carrying out a large number of measurements and their statistical processing. 11. It is necessary to study the generation of cracks in concrete and the methods of making structural joints. The Recommendations of the Congress. 1. The design engineers must draw the attention of the dam constructors to the necessity of establishing the suitable b u i l t - i n instrumentation during the construction. 2. The measurements of the b u i l t - i n instrumentation must be recorded periodically and frequently analysed by competent personnel and all unusual results of the observations must be investigated rapidly and carefully. The periodical and incidental visual observations of the whole region must complement the instrumentation observations on the dam. 3. The elaboration of general methods of testing and observations on dams of all types (earthfill, rockfill and concrete) must be organized and the information exchange between countries in this field must be accelerated. In earthfill dams with a core the crack formation might be reduced to a minimum by means of correct zoning and draining. In some countries cyclopean and boulder concrete masonry dams are of a special interest. 4. It is recommended that special studies should be carried out to establish the basic requirements regarding the safety of old or poorly constructed dams, and of dams built from inadequate materials or based on very new type of projects put in practice for the first time. 5. The programs of periodical inspection and evaluation of the condition of dams must be organized in all countries having dams since they represent a potential danger to the life of people and the safety of property. The data on eariier inspection investigations and investigations related to the construction period must be made available to the engineers in charge of the periodical inspection. It is recommended in that regard, to study the recently published UNESCO "Recommendations on Reservoirs" prepared by an international group of experts appointed by the general manager of UNESCO in accordance with the recommendation of the President of the International Commission on Large Dams. These recommendations enable the use by developing countries of the experience of industrialized countries in determining the procedures related to observations on reservoirs during their design, construction, and operation, and also on the deterioration of the related structures and the measures to e l i m i nate them. L I S T OF SOME OF THE PAPERS ON PROBLEM N O . 3 4 No. 5. D. McHeury and C. C. Oleson (USA), Pulse velocity measurements on concrete dams. No. 8. M . F . 130110, F. F. Bollo, and I. Rivet (France), Auscultation of concrete in situ. Investigations on the quality and possible anomalies of concrete. No. 13. C. D. Crosthwaite and I. K. Hunter (Gr. Britain), The deterioration of concrete dams: forty years of experience in North Wales.
'/4
V . V . STOL'NIKOV
No. 14. N. Smtterheim, I. A. P. Laurie, and N. Shand (South Africa), Deterioration of a multiple-arch dam as a result of excessive shrinkage of aggregate. No. 19. P. Cabanolis, A. Taillebot, P. Capponi, H. Cambefort, and A. Mager (France), Behavior and deterioration of dams. Maintenance and repair of dams. No. 21. S. S~illstr6m (Sweden), Plastic coating used for sealing old concrete dam. No. 22. G. Fristr0m and S. S~llstr0m (Sweden), Control and maintenance of concrete structures in existing dams in Sweden. No. 24. B. Mather (USA), Factors which affect the deterioration of concrete dams and measures for prevention of deterioration. No. 25. IL B. Jansen (USA), Behavior and deterioration of dams in California. No. 27. I. %lpsu (Turkey), Investigation of water losses at May reservoir. No. 28. IL Heggstad and R. Myran (Norway). Investigations on 132 Norweian concrete dams. Provisional resuits. No. 30. G. Willm and N. Beaujoint (France), The methods of supervising the behavior of dams used for hydroelectric energy by Elect~icit6 de France. Old problems and new solutions. No. 34. H. Sistonen (Finland), Montta and Seitakorva hydro-power plants. Sealing pervious layers underneath the earth dam and in the river banks by using the ICOS-Veder method. No. 35. B. Gilg (Switzerland), Long-term measures at the Mauvoisin arch darn. the course of time through measurements of the velocity of propagation of elastic waves. No. 49. A Rebaudi (Italy), Observation of the behavior of dams in the course of time through measurements of the velocity of propagation of elastic waves.