🏗️ Issue 08 — Erection Engineering & Temporary Works Dependency
The Tower That Needed a Schedule and Got a Dependency
Seven steel towers. A 173-metre giant among them. A release date they never controlled.
EE&HL Newsletter·August 2026·Issue 8·Free
The Millau Viaduct mid-construction, 29 June 2004 — the red temporary towers carrying the deck between permanent piers. Photo: Janekpfeifer, CC BY-SA 3.0, via Wikimedia Commons.← All Issues
Section 01✓ Free
The Project
173
metres
Tallest of seven temporary towers, standing through summer 2004
91
days
Deck closure to final stay-cable installation — dependency window
154
stays
Total cable-stays; last one fitted 27 Aug 2004, 12:04 pm
7
towers
Temporary supports erected across the Tarn Valley launch
(Millau Viaduct, Tarn Valley, France. Sources: Martin/Servant/Cremer/Virlogeux, Avignon Symposium 2004 [primary engineering account]; Bringer, Lecinq, Peltier, Buonomo & Servant, "Les haubans du viaduc de Millau," AFGC [primary]; Le Moniteur, 2002)
Seven steel towers rose out of the Tarn Valley from 2002, built to do one job: support the steel deck as it was pushed from opposite ends of the valley, across effective launch spans of about 171 metres, in repeated 600-millimetre hydraulic strokes. The tallest of the towers reached 173 metres, taller than most structures anyone would call a landmark in their own right.
They were never meant to be permanent. A 2002 execution plan stated their removal condition before a piece of deck had moved: once the deck was closed above the Tarn and the final cable-stays were installed and tensioned, the towers would come down.
That sentence did two things at once. It told the erection contractor exactly when the towers would stop earning their keep. And it made that date dependent on a separate specialist crew's own schedule — the stay-cable subcontractor working through the summer of 2004 while the erection contractor's own remaining pylons were still being trucked in and stood upright one at a time.
The deck halves reached each other on 28 May 2004, joined within the following two weeks. The last of 154 stay cables was fitted on 27 August 2004 — three months after closure, completing the installation phase of the towers' documented release condition.
💡 The Core Insight
A 173-metre structure with a release condition it didn't control. Not a load question. Not a wind question — both were closed out long before steel arrived on site. This was a sequence dependency: a structure whose exit depended on a specialist crew's own timeline, running in parallel, on its own clock. And that exposure sat with whoever was paying for the towers to remain standing — for as long as somebody else's specialist scope took to finish.
Foundations begin for the seven temporary, height-adjustable steel towers — built to reduce the effective launch span from 342 m to approximately 171 m at mid-span.
2002–2004
Deck launched from both abutments in 171 m stages, advancing in 600 mm hydraulic strokes over the permanent piers and temporary towers.
28 May 2004
Deck halves meet above the Tarn, joined within the following two weeks. The temporary towers' documented release condition begins its clock.
Jun–Aug 2004
Freyssinet, appointed by Eiffel in January 2002, installs the final 130 of 154 stay cables in line with the erection rate of the five remaining pylons.
27 Aug 2004
Last stay fitted, 12:04 pm. The stay-installation phase of the towers' release condition is complete — 91 days after deck closure.
17 Nov 2004
Load testing begins. Dismantling of the temporary towers continued through the autumn — 82 days beyond the release condition being met.
The bracket sits at the centre of the commercial pattern:
[days between deck closure and final cable tensioning] × [standing cost of a 173 m temporary structure, per day] = exposure
No monetary figures are named here — none are publicly sourced for this specific standing cost. But the variables are real, and they belonged to different parties: the erection contractor owned the towers. A specialist subcontractor, working to its own defined schedule, controlled when they could leave.
The Pattern & The Question
Why a temporary structure can have a perfectly defined removal condition and still have zero control over its own release date — and the one question to put in every temporary-works contract before it goes up.
Free to read in the full issue — sent straight to your inbox.
