Most information available on the Vajont rock slide is obtained from two detailed reports published by Muller (1964, 1968). It seems that from the very beginning, the chief engineer Carlo Semenza had some doubts about the stability of the left bank of the gorge. Several geological reports were submitted to him in 1958 and 1959. In 1959 he entrusted Muller with a detailed study of local conditions, which was carried out in collaboration with the geologists Eduardo Semenza and D. F. Giudici. Between February and November 1960, the water level in the reservoir rose from 580 m a.s.l.
to 650 m a.s.l. Movements of the rock were detected in September and November and when they reached about 3-5 cm a day it was decided to empty the reservoir down to 600 m a.s.l. Rock displacements rapidly stopped with the falling water level.
A first rock slide occurred on 4 November 1960 when 700 000 m3 of debris slid into the lake within about ten minutes, just upstream of the dam on the left bank. At the same time a long M-shaped fissure was detected high up on the slopes of Mount Toe clearly defining the area which was to slide.
Muller visited the area again in November 1960 and then after the major rock slide of October 1963.
According to Muller (1964), geological reports dating from 1958, con- firmed by seismic tests carried out a year later, concluded that the area was formed by in situ rock showing no sign of earlier movements. Dal Piaz described the rock near the gorge as being consolidated through cementation and he expected only minor superficial slides. In the Spring of 1959 Muller and E. Semenza reached a different conclusion, confirmed by detailed geological surveys. They felt that the rock mass on the left-hand side of the river was part of an old prehistoric slide which had come from Monte Toe.
Remains of this mass were found on the right bank of the Vajont, dis- cordantly stratified. The geology of the area is well known. At greater depth a Dogger formation is shaped like a concave saddle, the upper part of which is inclined at about 30° to 40°, the lower part being nearly horizontal. In this saddle, Malm and Cretaceous formations were resting in apparent equilibrium (fig. 14.1).
1300
Fig. 14.1 Idealized geological section through the left bank of the Vajont gorge.
(a) Vajont gorge; (Jb) north face of Mount Toe; (c) Pozza; (d) antithetic fractures;
(e) fracturing due to external rotation; (B) Dogger-malm formation (Oolitic lime- stone); (1) malm; (2) upper malm; (3-5) Lower Cretaceous; (6-8) Upper Cretaceous (after Broili, 1967).
Additional studies were carried out after the 1960 slide. Borings and seismic investigations followed. Some of the experts maintained their favourable diagnosis of the local rock conditions. Miiller and Semenza disagreed with them.
The rock masses were continuing to slide very slowly. Miiller's report analyses in detail the vertical and the horizontal displacements of the creeping rock masses and correlates water table levels in the rock (measured in boreholes) to the varying water levels in the reservoir basin. During 1962 the water in the reservoir rose steadily to level 700. In 1963, it slowly oscillated between levels 700 and 650 and back to 713, never reaching the dam crest (at 722-5). In the final days before the catastrophe, the velocity of the slide reached 20 to 30 cm a day.
Miiller, in his report says:
Surprisingly, no appreciable instability was observed throughout the period between 1961 and 1963, although during that time the water level twice reached an elevation of 100 m above the initial level of 1961 and 50 m above the level at the start of the creeping in 1960. During this time the effect of the buoyancy must rather have in- creased and reached its theoretical maximum value (one tonne per cubic metre) since in the meantime the degree of separation of the joint net and the permeability of the rock mass had increased, as can be seen from the gradual flattening of the mountain ground water level.
According to some remarks by Miiller, the water levels in the vertical holes (period from 1962 to 1963) varied with the reservoir levels, but *... no correlation could be established with the precipitation during the final
Period before 9 October 1963 405 months'. He thinks that pore pressure, water thrust and weakening of the shear strength in rock joints soaked with water were vital factors in this whole process. The final sentence in Miiller's report reads:
Such behaviour [last phase of the slide], connected with properties hardly investi- gated, and which can in a sense be regarded as thixotrophy, was completely un- expected, and it is the conviction of the author that the sliding could not possibly be foreseen by anybody in the form in which it actually took place and, in fact, nobody had foreseen or predicted it.
The engineers in charge of the Vajont dam had ordered hydraulic tests to be carried out on a very large model of the reservoir basin. They were based on geological findings. It was assumed that a large mass of rock would slide, being progressively accelerated so that the rockslide would occur in a few minutes. Tests carried out on these assumptions showed that there was no danger of the dam crest being overtopped, provided the reservoir level stayed below a determined safety level. When, at the beginning of October 1963, there was danger of an imminent rock slide, the water level in the reservoir was hastily lowered to this safety level.
The prevailing theory on the probable rock slide can be summarized as follows. All the information available to the Italian experts indicate that there is a progressive increase in velocity of the masses which are being accelerated from zero to the final speed, Such an acceleration would normally take several minutes. A time of about ten minutes was acceptable. Furthermore, observation of the creeping movement on the Monte Toe slopes indicated that the pressure of the rock caused rock masses on the edge of the gorge to be overthrown. It was thought that this type of movement would eventually fill the gorge and therefore bring the sliding movement to a stop.
The Italian engineers accepted the fact that the reservoir would be partially filled and maybe cut in two parts. They had already built a bypassing gallery on the right bank of the gorge which should have connected the upper part of the reservoir to the lower part, in case the falling rocks cut the reservoir completely in two. It is clearly established that everyone was fully aware that there was an immediate danger of a major rock slide. The actual rock slide was far larger in volume than expected by most of them, but its volume had been correctly predicted by Miiller and Semenza. Furthermore, it occurred with a violence and a suddenness which nobody had expected.
Miiller writes in his 1964 report:
. . . Only a spontaneous decrease in the interior resistances to movement would allow one to explain the fact that practically the entire potential energy of the slide mass was transferred without internal absorption into kinetic energy, and that the front of the sliding mass was pushed 400 m and 140 m up on the opposite slope, while moving over an 80 m wide gorge without falling down into it. Such behaviour of the sliding mass was beyond any possible expectation: nobody predicted it and the author believes that such behaviour was in no way predictable.