Some thoughts on dam safety10 June 1998
Rodrigo del Hoyo Fernández-Gago & José Alberto Herreras Espino consider some of the issues under discussion at the SpanCOLD meeting on dam safety
In the life of a dam it is usual to distinguish five phases, in accordance with the type of activities which, in relation to the dam referred to, are being developed. These five phases are the design, construction, first filling, operation and the ‘out of use’ state.
The five phases can be grouped together in three stages. The first would be the investment stage, which consists of the phases of design, construction and first filling, in which the dam is conceived, built and tested; a second stage of ‘rentability’, which can be considered as the one in which income and profits from the dam are collected, and which corresponds to the phase of operation, and a possible third and last stage, which will be the abandonment of the installations.
This last stage is not usual in large dams. We only have to remember, for example, the Proserpina and Cornalbo Dams, which were constructed during the Roman period in Spain, and which at the present time continue in operation. Dismantling has been carried out only in cases where an accident has occurred that has caused the failure of the dam. In such cases the river bed must be conditioned so the river can follow its course without problems, except when the dam has been reconstructed or a new dam constructed in its place. This last stage, which corresponds to the ‘out of operation’ phase can therefore be considered as exceptional.
Examples of the first two stages usually coexist all the time, but it can be said that in different periods of development of a country, one of them is preponderant. So, in the majority of developed countries, the investment stage has been the protagonist of the 1920s to 1970s.
The construction of dams varies from one country to another in accordance with their climatological and hydraulic characteristics. In the so-called green Europe the natural regulation of the rivers is some 40% of the average volume of flow — a consequence of climate, and topographical and geographical characteristic. In other countries, Spain for example, the natural regulation is only of the order of 10%, and it has been necessary to construct more than 1000 dams in order to obtain a regulation of 40% — equivalent to that available in a natural manner in other European countries.
For this reason, in Spain the construction of dams continued in the 1980s at a superior rate to that of other European countries, and it continues, although with lesser intensity, in the 1990s. At the present time Spain and the European countries, along with most of the developed countries, are in the second stage. This corresponds to the operation phase, in which the maintenance of the installations and their operation in order to recuperate the investments made are the principal operating requirements.
In the European Community there are a considerable number of large dams in operation — a number estimated in some thousands — of advanced ages of more than 30 or 40 years. This represents a very important heritage. In Spain itself, and taking into account the characteristics of its hydrology discussed above, the 1200 large dams now existing carry out a fundamental mission: to cover the water requirements of the population, along with industry, agriculture and energy development.
It has been said that the greater part of the European dams are more than 30 years old. This is true; and among them there are many that were built before the second world war, using structural, hydrological or geological models which today are considered to have been superseded.
In many of the dams which are currently in operation in Europe, the criteria related to safety, which today are applied to the dams both in their design and in their construction, have not been taken into account. That being so, we should remember, in agreement with the General Report of the Q-65 of the Vienna Congress, that many of the criteria relating to safety which are applied on designing a dam have, as their object, to cover a series of uncertainties existing in the design phase.
Many of these uncertainties may have disappeared in a dam with an extensive history in the operation phase. A dam with an operating history has of course shown that it performs adequately.
Starting from the basis that the conditions of safety should be the same for all dams, regardless of the period in which they were built, the necessity arises of evaluating the grade of safety of the dams in operation. In this evaluation we can take into account that for dams in operation information is available — which may be more or less complete, but is always important — and was obtained during the years that the dam has been in service, and would not be available for dams under design.
One problem that is often encountered in dams built some decades ago is that at the time they were designed, hydrological information was notably inferior to that of the present day. Methods for evaluating floods, for example, were based in many cases on simple empirical formulae.
When the time comes to estimate the maximum flood which could occur at the site of a dam, what is often forgotten is the historical ref-erences to floods that are made in countries with a long history, as occurs in Spain. These can be very useful in validating the statistical or determinist methods which are employed.
In Spain and other European countries, and principally in the romanised countries, there is in existence considerable information on floods which has been preserved over the length of their history, and which should be recovered.
This search for historical references on floods could commence in the first place in newspaper archives. In references in newspapers, many of which were already in existence more than 100 years ago, there is information, which in many cases is accompanied by graphical references, about the important floods of the last century or century and a half.
Following this, in the archives of the town councils and of churches, monasteries, and other organisations — that is to say, in the habitual sources of all historical investigation — it is possible to find additional data. This data is qualitative in nearly all cases, but it could be quantified with some acceptable limits. It should also be recalled that in many monuments there are marks of the water level reached on particular dates.
It would be very useful to develop, in Spain and perhaps also in other European countries, an investigation of this type. This would make it possible to validate the functions of the distribution of extreme floods most adequate for different regions.
One should remember that the countries that have developed determinist methods, using the concept of the permitted maximum flood (PMF) in order to establish the floods that a dam must support, are in general those of the New World, in which the historic references on floods cover a notably shorter period.
On the other hand, for dams in operation for one or more decades, very detailed quantitative data are available of the hydrographs that have appeared, which make it possible to establish a maximum hydrograph, validated by direct measurements.
On some occasions (especially in dams with a relatively large reservoir) to operate adequately, freeboards must be maintained during certain periods of the year and criteria must be established, as a function of the volume of inflow, for opening the gates. An alternative is to increase spillway capacity.
When dealing with concrete dams, it may be necessary to accept the possibility of overtopping, or, with old dams, adapting the dams in order to deal with extreme floods.
With the structural behaviour of a dam, it is often useful to extend the control and monitoring. This is despite the fact that in most cases no abnormalities have been observed during the operating period. Equipping the dam as far as possible with a modern monitoring system, allows the operator to go some way towards quantifying its behaviour.
In the last two decades there have been many old dams, generally concrete or masonry, in which monitors have been installed. Drains and piezometers have been added to relieve the uplifts and control them; pendulums have been installed, making use of existing wells or else drilling them for this purpose; and systems for topographical monitoring and devices for measuring the movements of fissures and joints have been added. This type of operation, in general not very costly, will improve analysis of the structure’s behaviour.
While extending the instrumentation and automatic monitoring, one thing that should never be forgotten is detailed visual inspections, to register anomalies in the structure and in the rock mass. This was almost the only means of control before modern monitoring systems were developed, and it continues to have great value in guaranteeing safety, both in old and new dams.
One problem that must be taken into account is the ageing of the materials of the dam. Concretes age and lose their resistant characteristics over time; for example via the alkali-aggregate reactions which were first detected in the 1940s. These are expansive reactions that have become clear after twenty years or more of the life of a dam, and which are caused by an alkali-silica-type reaction involving both the alkaline ions and the active silica, coming from more or less altered aggregates.
Much research work is under way to investigate this problem, particularly in countries where dams are in the ‘rentability’ stage. Research largely focuses on the reaction’s cause, which is influenced by the percolated volumes of water, the temperature and the tensional state among other factors, and looks for potential solutions.
In Spain, the legislation relating to the design, construction and exploitation of large dams — and consequently their safety — is at a point which could be referred to as a transition period. That is especially so because as it refers to dams in operation, which number more than 1100, there exist two groups of technical determining factors. These groups are legally different, and are applied to dams and reservoirs according to their ownership.
In effect, the Technical regulation on the safety of dams and reservoirs — valid since 1 April 1996 — is only applicable, for the moment, to state-owned dams. All other dams are governed by the very much older (1967) Instruction for the design, construction and operation of large dams. However, the Regulation also obliges all dams to be classified as a function of the potential risk of failure or incorrect operation or functioning. This classification should have been proposed by the proprietors and agreed by the Administration by the end of April 1998.
As is logical, since it is more modern, the Regulation introduces a series of obligations relative to safety. These obligations include periodical inspections by teams of independent technicians, and the availability of emergency plans which are obligatory for all the dams classified in the following categories:
•Category A — the failure or incorrect functioning of which can gravely affect population centres and essential services and lead to material or environmental damage.
•Category B — the failure or incorrect functioning of which can give rise to material or environmental damage or affect a reduced number of houses.
The recent completion of the classifications and requirement to draw up emergency plans will very probably speed up the extension of the Regulation to all dams in a relatively short time. This is desirable because — as well as homogenising the procedures — its prescriptions are, without doubt, very much closer to the actual technology.
Related ArticlesSpotlight on...Europe