Website dedicated to the disaster of Malpasset

Was it possible to avoid disaster?

It is almost impossible to answer yes to this question. The official conclusion of the court citing a "veritable trap of nature" is an indisputable fact. But could the trap be identified before the tragedy? A categorical negative answer is not scientific.


Above all, it is useful to clarify the nature of this trap, knowing that there was no eyewitness and that the bulk of exhibits, especially the trap itself, disappeared in the disaster. Only remains the footprint, enormous gaping wound and a small section of concrete stairs, relic on which many theories of experts have been drawn up.

50 years later, it is possible to make a concise description of the official thesis by cross-checking credible approaches: "The failure is due to the presence of a fault parallel to the dam dipping at 45 degrees below it to a depth of 15 to 17m from downstream to upstream and to a particular characteristic of the Malpasset rock consisting of a highly variable modification of his permeability under compression. The combination of these two factors plus a mechanism of uplift pressure has caused a true effect of hydraulic jack lifting the compressed rock. A last bad characteristic of the rock defined by a shift of its layers rolling over each other in certain conditions make worst the situation. As a result of  these dreadful combinations, several thousand m3 of ground literally exploded under pressure, depriving arch of its left bank support. The left bank thrust block was then found alone to withstand the water pressure. Not expected to contain such an increase of thrust the abutment moved back 2 meters, impairing the arch effect in the thin shell which was instantly broken on virtually its entire length".

The key word that comes to mind: geology.

It is true that it is only since Malpasset dam break that this discipline was placed on center stage of research firms and that the profession of engineer-geologist was created, but even so, was it not possible to understand the inability of ground to serve as reliable support for an arch dam? Why the work of an investigation committee formed in 1934 has not been taken into account in the preliminary study? Why Boson mines experts have not been consulted? Also not taken into account retrospectively, the geological report of the construction company of the motorway near the dam indicating the extremely slippery ground when subjected to bad weather. Many unanswered questions that leave doubtful on the sole responsibility of nature.

We are at the heart of the Malpasset case: lack of discipline and lack of communication on a background of absolute confidence in technology.

No explicit human responsibility has been upheld by the courts. It is nevertheless possible to formulate a number of general comments regarding compliance with the rules of the art in design and project management of dams based on verified facts. 

Insufficient of preliminary study is acknowledged. In court hearings, 4 experts from 11 maintained their explicit position on this issue. The others did not dispute this fact but claimed that further study would not identify the nature of the trap. Some of these specialists have argued that the supposed fault implied in the origin of the drama was not detectable but the doubt remains. Even if there was only one chance in 100 to find the fault, the use of 27 million francs provided for the preliminary study would be allowed instead of 5 really used, thus confirming the doubts about the appropriateness to construct a thin vaulted dam at this place. By creating a reservoir on a stream that has no water for most of the year, it was not necessary, as is commonly used to divert its course during construction. This is done by digging a tunnel to temporary drain the water through a lateral channel. Such a tunnel would probably intersected one of the major faults. Immediately after the disaster, tests were conducted to determine the variation of permeability of rock subject to compression. They found huge discrepancies, well above normal. Why these simple tests to perform, as we have been confirmed by those who have made its, have not previously included in the study of local geology? 

The geologist should be able to provide all required information to validate the well-founded of setting up a work at the expected place. He must accept his mission only if he has all the necessary means to hold on it well. George Corroy was a scholar well known in the region for his writings on Mediterranean geology. He was originally selected by Agricultural Engineering for a matter of proximity, long before the arrival of the engineer who met him only once, their communication was done by writing. 

Had the geologist a really field experience enabling him to focus its mission on the key points for choosing location of an arch dam? Probably not. We believe that despite his great qualities as a scholar, George Corroy did not have all the skills to prevail in this particular area. Note in this regard that he proposes in his final report the Sarrans dam as an example for the design of a sealing veil. This building is a gravity dam with has no common point with Malpasset. Moreover, the Malpasset issue is not a question of sealing veil, but the mechanical behavior of soil under the pressure of a thin arch dam. 

Recognition of insufficient research and preparation is confirmed, but there is no responsible because it is not provable that the absence of these deficiencies involves explicitly the absence of disaster. 

The engineer must ensure that he has all the guarantees of very good geological quality of the soil on which he will construct his dam. He must ensure that the work leading to the conclusions is sufficient and if this is unavoidable, he should not hesitate to reappraise the choice of thin arch dam in favor of another type of structure. 

André Coyne is considered as the grand master of arch dams and this remains true despite Malpasset. He was really revered in his profession. According to witnesses, confirmed during the trials, André Coyne decided everything. He was feared. During meetings, no one dared speak as the character was so impressive in his image and his keenness. In the preface of several technical books dealing with dams, we can read "Master of us all" In a book telling the history of Electicity of France, the author Pierre Gerard said: "At least it was appropriate even so not to play too self-confident, temptation which lurks ambitous managers who have never met setbacks and no longer used to being challenged. In any profession besides, the reign of any pope is really desirable.". 

or: "A trial took place. It must be said that in most cases, what is said at trial, referring to the legal code and under the influence of media opinion, is simply ridiculous. The guilty party? It is human intelligence that has not been able to analyze one aspect of the stability of dams, thinking afterwards, however, was within reach. Too many experts argued that the problem of under-pressure, so general in the dams does not appear in the foundations of arch dams, and even then it was considered as their big advantage. This is the collective intelligence of specialists in the profession that had to appear before the court.".

We must study the environment of the time to understand the importance of such characters: the effective creation of technocrats so decried in the '70s by sociologists. The watchword of postwar was "rebuild" but also "build" because something new, innovative was wanted. The future belonged to the ambitious, those who could demonstrate initiative. The engineers were considered far above the others and the dam builders were with coal miners the pioneers of reconstruction and national independence. The high school of Highways, founded in 1747, which André Coyne came from represented the peak of the hierarchy.

The presentation of the first General Plan Commissioner is explicit to have one's own idea about the ambitions of the time:

  • To go up again in 1948 the production level of 1938;
  • To reach in 1949 the level of 1929 then 25% more in 1950;
  •  These goals were not achieved in 1950 but exceeded in 1952.

The technocrats of this period had very extensive powers which placing them in the heart of decisions. As a key witness of this corporation told us : "We feel a bit like the masters of the world.". André Coyne was a kind of "master of masters of the world" He had the opportunity to impose whatever he wanted. Nobody would have objected. Has he done everything in his power to have all the information necessary to validate the project, particularly regarding the reliability of soil support? We must face the evidence, the answer is no. his vaults were and remain very strong, but at Malpasset the rock would must be much more observed because it is highly specific and extremely heterogeneous. It is André Coyne who chose the final place, 200 meters downstream from the one recommended by the geologist. It is simplistic to say today that the dam would probably resisted if it had been built on the original location. That's the way. 

One more time, and it is especially significant in the Malpasset case, this major decision was taken following a simple correspondence between the engineer and geologist. 

An engineer of Highways Department told us that as soon as the disaster was announced to him, he was not able to accept the dam break. This option was not within the framework of its technical expertise. 

André Coyne was no longer in this world during the trial. He was probably the main accused. Moreover he immediately told his partners and in particular its customer, the Agricultural Engineering, he was solely responsible. His declaration was also made to the investigating judge who came to meet him shortly before his death. The engineer is the only major player in the Malpasset case who has explicitly acknowledged its responsibility. The trial took place without him and it appears clear from a reading of archival documents, that this has grealty contributed to the simplification of proceedings. 

The cultural distance and lack of effective communication between the geologist and the engineer are the two factors appearing significant deficiencies in the design and construction of the dam. 

The contractor must ensure that all necessary resources for the successful implementation of the project entrusted to him are well combined. He should also act as a conductor and constantly ensure that partners who work for him and with him form a cohesive team, knowing to communicate together and systematically submitting to regular review meetings. It is clear that given this definition of its mission, the Agricultural Engineering in the person of its chief engineer Jacques Dargeou failed. And as such he potentially beared responsibility that was not recognized at trial because he was not master of this decisions. Who then should be charged? 

The context of the time has to be viewed in order to understand the situation. The Agricultural Engineering was the poor relation of technocracy which paled into insignificance beside the prestigious Highways Department. Regarding research and preparation works, it had no serious ability to judge their importance. So, it relied on the undoubted expertise of André Coyne. No one can bear it a grudge for this decision, but had it not opportunity to impose the mobilization of the initial budget of 27 million? Had it not possibility to ask the engineer to intercede on his behalf? 

The biggest fault of the engineer of the Agricultural Engineering is probably not be able to impose the geologist and the consulting firm to work as team. 

A last point but not the least concerns the project management. It is the acceptance of receipt of the dam while it had not been brought to full load. The first complete filling of a dam is an essential step to ensure its perfect response to the maximum thrust that it will endure. Again, the supervisor was satisfied with the garantee provided by the engineer based solely on his experience. 

Despite the exceptional bad weather of November 1959, how can we affirm the unpredictability of the sudden surge of water when such events were common and well known in the region? Who defined the hydrological parameters that led to the design of the drain valve? Despite its flowmetering consistent with current standards, would it not have been oversized to accommodate the sudden and rapid changes of the flow of Reyran stream knowing that the volume of water security between the exploitation of maximum level was relatively low? 

It reported the words of André Coyne after the disaster: "I crossed the Zambezi River and have an accident on a runlet ...." The Zambezi despite its pharaonic dimensions is probably more predictable than the Reyran. 

Another factor often mentioned in the writings on Malpasset, it is the blasting done during construction of Esterel-Côte d'Azur motorway. 

An extensive investigation conducted with the assistance of engineers, geologists and witnesses leads us to believe now, after long hesitation, that this event has played a major role in the disaster, at least as an amplifier of the trap. 

Recall the facts: in December 1956 began the construction of Fréjus to Cagnes-sur-Mer link, part of the Esterel-Côte d'Azur motorway. 

The building site moved into the houses of old mines of Boson previously occupied by workers of the dam. Some of them have resumed service for the motorway. 

In November 1957, intense trigger of blasting is carried out near the dam. In fact, that portion of the motorway is in high relief and the equipment of the company does not allowed to continue construction of the road without solid rock excavation. The construction company, whose expertise was largely regional, had participated actively in the construction of the Canal du Bas-Rhône Languedoc in a little steep. It had only a limited equipment to tackle the ground of the valley of Reyran. To continue the progress in respect of specifications, the company had massive use of explosives and appealed, according to several technical witnesses, a specialized team from Poland. Probably colliers but unfortunately we have very little information on this subject. 

The blasting campaign has, according to witnesses presents on the site, lasted several weeks. The guardian Ferro and the newly appointed engineer of Agricultural Engineering were deeply worried. The violence of the shock wave was such that the apparatus used to move the valve train of the dam violently hit the facing of the arch. Following repeated complaints from the guardian, a team was constituted, composed of representatives of the construction company, engineers of the motorway and members of Agricultural Engineering. The conclusion was reassuring: the dam was designed to withstand earthquakes of low intensity. 

We have good reason to consider this conclusion as very optimistic and here again a little precipitate. It was necessary to accurately study the impact of blast on the rock and not on the dam. The theory that the dam itself could act as a "hammer" capable, under the effects of repeated blasting, to cause a real "punching" of the rock immediately in contact with the foundations of the left bank is not be ruled out. Moreover, it is possible to consider, based on the testimony of engineers on the scene when facts appeared, that the dosages of powder and the conditions of implementation of blasts were different from those indicated in the official reports. These violent shock waves repeated may have the effect of facilitating the triggering of the trap. 

The technical supervision of the dam was reduced to a minimum. The very thin structure had no measuring equipment of dynamic deformations as it existed on similar works. It was apparently not seen fit to install one and in any event, there would not have qualified staff on hand to interpret results. This is not necessarily critical if an effective and continuous monitoring of deformations was put in place. To the discharge of the engineer of Agricultural Engineering, he has asked very quickly after the first water filling the drawing up of a monitoring and maintenance plan. He will obtain an official response only four years later. However, he will still conduct annual campaigns of deformation measurements of the vault. These changes are inevitable as a dam "lives" and "moves" until to take its final position. The important thing is the type and location of deformations. 

How to explain these delays outside funding problems? But has the Agricultural Engineering used every means at its disposal to achieve his ends? Was André Coyne really informed? And what was his position? We do not know. 

The funds have rapidly declined after the dam construction. Indeed, the funding expected from the Ministry of Defense (for water supply of military camps) and Ministry of Interior (for water supply of cities) were not obtained. This is the main reason why irrigation was never operational. The documents reveal that the mother branch, essential part of the water distribution system was temporarily approved during the year 1959. 

To return to the measurement campaigns, the last one was conducted in July 1959. It showed deformations, including one of 17-mm which merited special attention because of its location at the foot of the vault and taking into account that the measurement was made in the presence of a water level below the maximum . We must thus consider an increase of over 20% of the force applied to the roof when the maximum height was reached on day of drama. The observation of the W-shaped line of deformations shows that the arch was moved in a non-homogeneous way and it is extremely important. The foot of the dam was moving abnormally from the summer 59. Had the great flaw discovered at the foot of the dam after the disaster not occurred, so long before the tragic night? Guardian Ferro, who watched "his dam" every day often said to his interlocutor of the Agricultural Engineering, "I feel that the curve is no longer the same." It is the subjective opinion of a layman, which has probably a lot of sense. 

The results were communicated to the consulting firm four months after taking measurements ie very shortly before the disaster. These results have clearly not been really taken into account. Complex calculations needed to establish the results was claimed at trial, which is not incorrect given the limited technical resources available to companies at the time in France. But it is also recognized by experts that it was possible to quickly transmit raw results to the consulting firm, enough significant to make decisions. The contract with the measuring company did not include the use of methods less precise but faster and did not mention any delay for submission of results, so it did not provide the systematic delivery of information to the consulting firm without passing by the Agricultural Engineering. Thus, the court therefore found no charge against the measuring company because it was not subject to a contractual obligation to observe a maximum delay for transmission of results. 

Shortly before the disaster, the new head of the Agricultural Engineering, Leon Nourrit, in charge of surveillance, observed significant seepage below the dam, which accentuated quickly becoming real sources as the water plan gave up. He also noticed cracks in the protective carpet at the foot of the dam, additional evidence of its displacement. But the carpet protection, exhibit, disappeared completely during the breaking of the dam. The engineer called the companies involved in dam construction for a meeting on site for December 7, unfortunately, it was 5 days after the tragedy... 

In the three days preceding the disaster, the Agricultural Engineering asked the Highways Department permission for opening the valve while level 96 was not yet reached ie 4 meters before overflow. The answer was negative on the fact that release of water had been detrimental to the motorway bridge under construction. On the first of December, the maximum operating level was reached but the system for automatically opening the valve did not work. It should be set at the maximum level, which as everybody now knows never happened before the break. 

Was it possible to reduce the devastating effects of the wave? 

It is very likely. Firstly, we must reconsider the design of the dam. It was a very thin vault type. For an arch dam, the failure of support must surely lead to its destruction, but it is reasonable to assume that if its thickness was more important, its destruction would have been less and certainly more progressive. Moreover, the ground has dropped in the area where the thickness was reduced significantly. 

The less powerful wave would probably saved many lives. 

If the planned operation had taken place, it is highly likely that available resources would have been increased, but it is not certain. The water never arrived at the irrigation places for lack of money. The dam remained passive, with no other purpose than to provide a magnificent stretch of water. The role of the guardian was mainly dedicated to monitoring the building and for some routine operations. In the last year of life of the dam, Andre Ferro was virtually the only one continuously on site. Engineer Jacques Dargeou and his second were been replaced. 

Warning plan has not been drawn, nor immediate security zone of flooding wave propagation. The full destruction of the small village of Boson was inevitable. A car with a loudspeaker went up the valley of Reyran to inform farmers about the imminent release of water. There is no mention of its passage in Frejus. A simple information sheet was placed on a panel in the town hall. The possibility of a real threat given sudden increase of additional pressure on the roof of the dam was not considered and consequently, no information about this eventuality has been communicated to public. 

Guardian Andre Ferro lived in a temporary house that become definitive by lack of credit. Construction plans were drawed up, however, by an architect and approved by the General Council. The guardian narrowly escaped the disaster. He was absolutely unable to do anything to give warning. 

No human responsibilities and yet valuable lessons ... 

    * No dam was rebuilt in the valley of Reyran. An inter-union was formed with the township of Le Muy, Puget sur Argens and Roquebrune-sur-Argens for water supply from the Argens river. A treatment plant water was built in Muy and 17 km of pipes have been installed. Today, the syndicat includes 8 municipalities; factories Muy of Gargalon and Fourmel purify water obtained from several catchments in the nearby streams (Siagnole, Argens, Fournel and dam of St. Cassian) .

    * The Reyran was channeled after the disaster by concrete banks. 

    * The science of dams has changed tremendously after the disaster. There is a "before" and  an "after" Malpasset. The engineers were heavily questioned. 

    * Virtually all arch dams in the world were subjected to exceptional measures campains leading sometimes to modifications to take account of the discoveries made at Malpasset. Electricity of France in particular has greatly contributed to the logistics provision for expertise works. 

    * Water level of many dams has been lowered after the disaster and  in some case has not been raised. Dams have been emptied and / or modified. Projects have been abandoned (eg St. Cassien in France which would serve as inputs to the retention Malpasset) to be replaced by gravity dams. 

    * In 1964 the first Civil Security Unit was established in Brignoles with conscientious objectors supervised by members of mobile security. Two other courses of this type are now respectively located in Nogent-le-Rotrou and Corte. 

    * In 1966, as a direct result of Malpasset Dam break and demonstrating the lack of project management, it was created a Permanent Technical Committee on Dams (CTPB) whose mission is as follows: "The CTPB is required to give its opinion on the prior -projects and implementation projects of dams on the competence, powers and capacity of the Construction Manager and on the technical problems that may arise during construction and are likely to cause a substantial change of execution projects. Whatever the project manager is, the Permanent Technical Committee on Dams must be consulted by the Minister concerned on the draft proposals and implementing projects of dams with a height of at least twenty meters above the lowest point of natural ground". It is surprising that the mission of CTPB describes exactly what has failed in Malpasset ... 

    * In 1967 the French Committee of Rock Mechanics is established. This discipline became an essential science that dam designers regularly use. All research and design firms have a geology department and the profession of engineer-geologist has been specifically created. 

    * Decree No. 68-450 of May 16, 1968 provides for contingency plans set up by the prefects for the evacuation of the population. 

    * Circular No. 70/15 of August 14, 1970 cover inspection and supervision of dams involving public security and complements the laws prior to Malpasset dam break.