Preliminary studies


  • 1987: Establishment of the first plans by CETE of Aix-en-Provence;
  • 19 October 1991: Choice of the high crossing of the Tarn River by a structure of around 2,500 m (8,200 ft);
  • 1993–1994: Separate consultations with seven architects and eight structural engineers;
  • 1995–1996: Second definition study with five associated architect groups and structural engineers;
  • 10 January 1995: Declaration of public interest;
  • 9 July 1996: The jury decides in favour of a cable-stayed design with multiple spans, as proposed by the Sogelerg consortium (Michel Virlogeux and Norman Foster);
  • 20 May 1998: Decision to proceed by grant of contract;
  • 8 June 2000: Launch of the contest for the construction contract, open to four consortia;
  • March 2001: Eiffage establishes the subsidiary Compagnie Eiffage du Viaduc de Millau (CEVM) and is declared winner of the contest and awarded the prime contract;
  • August 2001: Ratification by the State Council on the project, allotting the contract to Eiffage by ministerial decree;
  • 8 October 2001: Decree formally approving the grant of contract.

Possible routes

Routes of the four projects of the A75 autoroute around Millau

Routes of the four projects of the A75 autoroute around MillauIn initial studies, four options were examined:

  1. An option called Great Eastern (grand Est) ( yellow route ) passing east of Millau and crossing the valleys of the Tarn and Dourbie on two very high and long bridges (spans of 800 m/2,600 ft and 1,000 m/3,300 ft) whose construction was acknowledged to be problematic. This option would have allowed access to Millau only from the Larzac plateau using the long and tortuous descent from La Cavalerie. Although this option was shorter and better suited to the through traffic, it was abandoned because it did not serve the needs of Millau and its area satisfactorily.
  2. An option called the Great Western (grand Ouest) ( black route ), longer than the eastern option by 12 km (7.5 mi), following the Cernon valley. Technically easier (requiring four viaducts), this solution was judged to have negative impacts on the environment, in particular on the picturesque villages of Peyre and Saint-Georges-de-Luzençon. More expensive than the preceding option, and serving the region badly, this option was also abandoned.
  3. An option called near RN9 (proche de la RN9) ( red route ), would have served the town of Millau well, but presented technical difficulties and would have had a strong impact on existing or planned structures. This option was also abandoned.
  4. An option called intermediate (médiane), west of Millau ( blue route ) had the blessing of local opinion, but presented geological difficulties, notably on the question of crossing the valley of the Tarn. Expert investigation concluded that these obstacles were not insurmountable.

The fourth option was selected by the ministerial decree on 28 June 1989. It encompassed two possibilities:

  • the high solution, envisaging a 2,500 m (8,200 ft) viaduct more than 200 m (660 ft) above the river;
  • the low solution, descending into the valley and crossing the river on a 200 m (660 ft) bridge, then a viaduct of 2,300 m (7,500 ft) extended by a tunnel on the Larzac side.

After long construction studies by the Ministry of Public Works, the low solution was abandoned because it would have intersected the water table, had a negative impact on the town, cost more, and lengthened the driving distance.

The choice of the “high” solution was decided by ministerial decree on October 29 1991.

After the choice of the high viaduct, five teams of architects and researchers worked on a technical solution. The concept and design for the bridge was devised by French designer Michel Virlogeux. He worked with the Dutch engineering firm ARCADIS, responsible for the structural engineering of the bridge.

Choosing the definitive route

Satellite image of the route before construction of the bridge.

The "high solution" required the construction of a 2,500 m (8,200 ft) long viaduct. Obviously, this would be the crown jewel of the entire A75 autoroute project. From 1991 to 1993, the structures division of Sétra, directed by Michel Virlogeux, carried out preliminary studies and examined the feasibility of a single structure spanning the valley. Taking into account technical, architectural and financial issues, the Administration of Roads then opened the question for competition between structural engineers and architects to widen the search for realistic designs. By July 1993, 17 structural engineers and 38 architects presented themselves as candidates for the preliminary studies. With the assistance of a multidisciplinary commission, the Administration of Roads selected 8 structural engineers for technical study and 7 architects for the architectural study.

Choice of technical design

Simultaneously, a school of international experts representing a wide spectrum of expertise (technical, architectural and landscape), chaired by Jean-François Coste, was established to clarify the choices which had to be made. In February 1994, on the basis of proposals of the architects and structural engineers, and with support of the school of experts, five general designs were identified.

The competition was relaunched: five combinations of architects and structural engineers, drawn from the best candidates of the first phase, were formed to each conduct in-depth studies of one of the general designs. On 15 July 1996, Bernard Pons, minister of Public Works, announced the decision of the jury constituted of elected artists and experts and chaired by the director of highways, Christian Leyrit at the time. The solution of a cable-stayed bridge, presented by the group of research bureaus Sogelerg, Europe Etudes Gecti and Serf and the group Foster + Partners was declared the best.

Detailed studies were carried out by the successful consortium, steered by the highways authority until mid-1998. After wind tunnel tests, the shape of the road deck was altered and detailed corrections were made to the design of the pylons. When the details were eventually finalised, the whole design was approved in late 1998.


Once the Ministry of Public Works had taken the decision to offer the construction and operation of the viaduct as a grant of contract, an international call for tenders was issued in 1999. Four consortia tendered:

  • Compagnie Eiffage du Viaduc de Millau (CEVM), led by Eiffage
  • a consortium led by the Spanish company Dragados, with Skanska (Sweden), and Bec (France)
  • Société du Viaduc de Millau, including the French companies ASF, Egis, GTM, Bouygues Travaux Publics, SGE, CDC Projets, Tofinso and the Italian company Autostrade
  • a consortium led by Générale Routière, with Via GTI (France) and Cintra, Nesco, Acciona et Ferrovial Agroman (Spain).

The Compagnie Eiffage du Viaduc de Millau, working with the architect Sir Norman Foster, was successful in obtaining the tender. The fact that the government had already taken the design work to an advanced stage meant that the technical uncertainties were considerably reduced. A further advantage was that it made the process of negotiating the contract easier, reducing public expense and speeding up construction, while minimising such design work as remained for the contractor.

All the member companies of the Eiffage group had some role in the construction work. The construction consortium was made up of the Eiffage TP company for the concrete part, the Eiffel company for the steel roadway (Gustave Eiffel built the Garabit viaduct in 1884, a railway bridge in the neighboring Cantal département), and the ENERPAC company for the roadway's hydraulic supports. The engineering group Setec has authority in the project, with SNCF engineering having partial control.

Appia was responsible for the job of the bituminous coating on the bridge deck, and Forclum for electrical installations. Management was handled by Eiffage Concessions.

The only other business that had a notable role on the building site was Freyssinet, a subsidiary of the Vinci Group specialising in prestressing, which was entrusted with installing the cable stays and putting them under tension, while the prestress division of Eiffage was responsible for prestressing the pillar heads.

The steel deck and the hydraulic action of the deck (the technical solution that had been successful in the competition for the design of the metallic moving parts) were designed by the engineering firm Greisch (BEG) from Liège. They carried out the general calculations and the resistance calculations for winds of up to 225 km/h (140 mph).

The sliding shutter technology for the bridge piers came from PERI.

Costs and resources

The bridge's construction cost up to €394 million, with a toll plaza 6 km (3.7 mi) north of the viaduct costing an additional €20 million. The builders, Eiffage, financed the construction in return for a concession to collect the tolls for 75 years, until 2080. However, if the concession is very profitable, the French government can assume control of the bridge in 2044.

The project required about 127,000 cubic metres (166,000 cu yd) of concrete, 19,000 metric tons (21,000 short tons) of steel for the reinforced concrete, and 5,000 metric tons (5,500 short tons) of pre-stressed steel for the cables and shrouds. The builder claims that the bridge's lifetime will be at least 120 years.


Numerous organizations opposed the project, including the WWF, France Nature Environnement, the national federation of motorway users, and Environmental Action. Opponents put forward several arguments:

  • The westernmost route would be better, longer by three kilometres but a third of the cost with its three more conventional structures.
  • The objective of the viaduct would not be achieved; because of the toll, the viaduct would be little used and the project would not solve Millau's congestion problems.
  • The project would never break even; toll income would never amortise the initial investment and the contractor would have to be supported by subsidies.
  • The technical difficulties were too great and the bridge would be dangerous and unsustainable; the pylons, sitting on the shale of the Tarn Valley, would not support the structure adequately.
  • The viaduct represented a detour, reducing the number of visitors passing through Millau and slowing its economy.