Lecturer at Sciences Po Paris, EM Lyon, IEP Lyon, ILERI, HEIP · President of the France-Formosa Association

Taiwan imports 97% of its primary energy. Its strategic reserves of liquefied natural gas cover between 11 and 14 days of consumption. Its last nuclear reactor was shut down in May 2025. And at the same time, TSMC’s electricity consumption — the company that manufactures over 90% of the world’s most advanced chips — could triple by 2030. The scale of the problem is clear.

An electricity mix on fossil life support

Taiwan’s energy portrait is one of near-total dependence on imports. In 2024, liquefied natural gas accounted for 42.7% of electricity generation and coal for 33.5%, all of it imported. Renewables plateaued at 11.6%. The government’s “532” target — 50% gas, 30% coal, 20% renewables — has already been postponed twice and will not be met in 2026.

This dependence on LNG, which enjoys rare bipartisan consensus between the DPP and the KMT, is a rational choice in peacetime: gas is cleaner than coal, less costly to deploy than renewables, more popular than nuclear. But it’s a risky strategic bet once the geopolitical variable is factored in. Supply transits through the Strait of Hormuz, the Strait of Malacca, and the South China Sea — all chokepoints. The island’s three LNG terminals are concentrated on the west coast, facing the strait, within direct range of People’s Liberation Army missiles.

Nuclear power: a democratic choice with strategic consequences

The closure of the Maanshan plant on May 17, 2025, marked Taiwan’s entry into its first summer without nuclear power in four decades. The referendum history on this subject is worth examining: in 2018, a clear majority voted to maintain the nuclear plants, well beyond the required threshold. The government proceeded with closure anyway. In August 2025, a new referendum on reopening failed to meet the participation threshold, although an overwhelming majority of those who voted supported reopening.

This creates a paradoxical situation: public opinion largely favorable to nuclear power, but an institutional and political framework that prevents its restart. The DPP, long anti-nuclear as a political identity marker, made the nuclear exit an ideological touchstone. The KMT, favorable to maintaining nuclear, failed to capitalize on referendum results. The outcome: a democratic deadlock on a national security issue.

Renewables: ambition constrained by geography

Taiwan has bet big on offshore wind. The targets are ambitious: 5.6 GW installed by 2025, 20 GW by 2030. But reality is more nuanced. Typhoons impose considerable technical constraints on turbines. The Taiwan Strait, despite its powerful winds, presents some of the world’s most challenging maritime conditions for installation and maintenance.

Solar, meanwhile, faces an available surface problem. Taiwan is a mountainous island of 36,000 km², two-thirds of which is uninhabited. Agricultural land is protected, exploitable rooftops are limited. Solar yields, decent in the south, drop significantly in the industrialized north.

TSMC: the world’s energy Achilles’ heel

The TSMC dimension transforms Taiwan’s energy question into a global issue. The company already consumes approximately 6% of the island’s total electricity. With the ramp-up of 3nm and 2nm chips, this consumption could triple by 2030. TSMC has made ambitious carbon neutrality commitments, but they rely essentially on purchasing renewable energy certificates — an accounting strategy that doesn’t change the physical reality of the electricity mix.

If Taiwan experiences a major electricity shortage, it’s not just the local economy that suffers: it’s the entire global semiconductor production chain that wobbles. TSMC’s diversification plans — plants in Arizona, Japan, Germany — represent only a fraction of installed capacity in Taiwan.

The blockade scenario: when energy becomes a weapon

The most acute risk isn’t invasion, but blockade. China could, without firing a single shot, strangle Taiwan by intercepting LNG tankers in the strait. With 11 to 14 days of LNG reserves, the island would be forced into massive load-shedding in less than two weeks. Coal plants, which have larger stocks (approximately 30 days), would temporarily take over — but at the cost of unbearable pollution and insufficient capacity.

This scenario is taken seriously by American and Taiwanese military planners. It partly explains the recent decision to build new LNG storage capacity, but construction timelines (5 to 7 years) make these projects useless against a short-term crisis.

Toward an energy resilience doctrine

Taiwan’s energy security can no longer be thought of solely in terms of optimal electricity mix. It must integrate a strategic resilience dimension: diversification of supply sources, decentralization of production, massive storage, and probably — however unpopular it may be for some — an honest reassessment of the nuclear option.

The question is not whether Taiwan can afford nuclear power. It’s whether it can afford to do without it in a context where every day of guaranteed electricity production is a day of preserved sovereignty.