The Limiting Factor Is No Longer Production

In heavy industry as in energy, the link that breaks first is rarely the plant or the well. It is the mandatory passage: the road, the railway, the strait, the power line. The articles of recent days, once placed back on the map, outline several of these tension points. They do not predict rupture, but they indicate where flows are concentrating too heavily.

Guinea: A Second Corridor for Bauxite That Is Overflowing

The clearest case comes from Kamsar. Nimba Mining Company, created in 2025, signed an agreement on November 4, 2026, opening access to the Multi-User Mining Road, a 56-kilometer infrastructure linking the southern part of its Tinguilinta permit to the port facilities at Dapilon. This axis, already used by SMB, CDM, KBM, Dynamic Mining and Ashapura, becomes an alternative to the rail corridor to Kamsar.

The logic is industrial: production must reach 9 million tonnes by late November, then 10 million by December 31, 2026. Its managing director explicitly speaks of "lifting a logistical bottleneck." Pooling an existing network avoids building new infrastructure, subject to access conditions and available capacity. But the 10-million-tonne threshold applies to bauxite extracted, not volumes shipped: the whole issue is whether the new corridor suffices to move what comes out of the ground.

Hormuz: The Lesson of the Single Passage

At the other end of the chain, the Strait of Hormuz illustrates the opposite risk. About 20% of global oil supplies transit through it, and the passage has been closed or nearly closed for much of 2026. The consequences show up in LNG: the March strikes removed 17% of Qatar's export capacity, and force majeure on deliveries to Asia and Europe has been extended to end-December 2026. Qatar's energy minister recalls that the strait cannot be bypassed by pipeline without rebuilding redundant facilities, which would have "no economic sense."

Traffic data, however, are contradictory. A recent analysis puts maritime traffic at 82-92% of pre-war levels by late September, with Middle Eastern oil exports around 24 million barrels per day thanks to Saudi and Emirati pipelines. Yet Qatar is moving only "very few" cargoes. Claims of strikes on tankers, for their part, are not independently verified. These figures should be read as fragile traffic, not as a return to normal.

Diversification as a Signal

The reaction of the players is instructive. Qatar has signed with Cheniere for up to 4 million tonnes per year from 2031, and with Woodside for 3 million tonnes per year from 2030, which would make it the largest buyer of US LNG. For its part, TotalEnergies announced a preliminary agreement to deliver 900,000 tonnes per year for fifteen years, starting in 2028, to the Rudong terminal in Jiangsu province, totaling 13.5 million tonnes. The province would rank second in China for GDP and gas consumption. The agreement remains conditional: neither price nor origin of cargoes is specified.

These contracts outline routes that no longer pass through the same point. They are staggered from 2028 to 2031: they respond to a long-term vulnerability, not to current tension.

The Power Grid, Bottleneck of the Decade

Another bottleneck is taking shape, less visible because it is immaterial. A note published on November 4 on the Banque de France website estimates that annual electricity demand from data centers could grow by 45 terawatt-hours in Europe by 2030. Frankfurt, London, Amsterdam, Paris and Dublin concentrate about three-quarters of computing capacity, 118 centers are under construction in major hubs, and in Dublin they would already account for 20% of national electricity consumption. The authors propose varying electricity prices by location, as in Sweden, divided into four zones since 2011, to steer projects toward territories where grid and generation are available. Electricity can represent up to 60% of the operating costs of these facilities; they estimate at 54% the share of consumption potentially relocatable. This is a proposal, not a rule in force, and the authors themselves highlight its limits (latency, sovereignty, water, local acceptability).

The mechanism is identical: a uniform national price masks local saturation, and the constraint appears in the form of connection delays in a few metropolitan areas.

Urban Signals: Fragility of People Flows

At the territorial scale, two signals point the same way. In Île-de-France, forecasts for the November 5 strike announce a normal metro but RER D at one train in three on several branches, RER C and RER B north at one train in two. In Guadeloupe, businesses in the Jarry zone, overwhelmingly micro-enterprises and SMEs, have described for three weeks falling revenues due to the blocking of the Voie Verte, with traffic jams and lost time. These episodes are local and temporary, but they show how a single axis conditions the activity of an entire zone.

What the Dossier Does Not Say

Several limitations stand out. No traffic, inventory or lead-time data allows these tensions to be measured, and no open data accompanies the dossier. The construction sector, part of the initial angle, appears in no usable signal. Finally, the location of each tension is inferred from the content of the articles, which remains an analyst's reading.

Conclusion: Watch the Passages, Not the Plants

The common thread is simple. Where a growing volume uses a single corridor, transit capacity takes precedence over production capacity, and players end up opening alternative routes or rethinking the price signal. Guinea is doing so in a few weeks, Qatar and TotalEnergies over horizons from 2028 to 2031, and the reflection on data center electricity is only beginning. To anticipate the next bottlenecks, one must therefore track single crossing points and bypass announcements. Their precise timeline and scale remain to be established with data the dossier does not contain.

This article was written with the assistance of artificial intelligence and verified by our editorial team.