Unintended Consequences — Episode Brief

Topic: Switzerland's Nuclear Phase-Out and the Coal Rebound

Segment 1 — The Cold Open

On the afternoon of December 20, 2019, the Mühleberg nuclear plant on the Aare River west of Bern disconnected from the Swiss grid for the last time. The 373-megawatt reactor had run since 1972; its owner, BKW, shut it on schedule, and Switzerland’s own power stations looked, on paper, a little cleaner by nightfall. That winter, when the sun dropped behind the Jura and electric heat came on across the Mittelland, a share of the missing megawatts arrived as imports from the north — and a measurable share of those imports were generated in German coal and lignite plants. This was designed to take a cautious, climate-conscious country off nuclear power. Instead, it shifted part of Switzerland’s winter carbon footprint across the Rhine.

Segment 2 — The Good Intention

The decision did not begin as a crusade against carbon accounting. It began as a safety judgment after March 11, 2011, when the Fukushima Daiichi accident forced governments everywhere to look again at aging reactors. Switzerland’s fleet was small and well run, but it was also old: Beznau 1, on the Aare in canton Aargau, had been connected in 1969. On May 25, 2011, the Federal Council, with Doris Leuthard at the energy department, announced that no new nuclear plants would be built and that the existing five would not be replaced when they reached the end of their safe lives. That stance hardened into Energy Strategy 2050 — a package of efficiency targets, support for new renewables, and a legal ban on new nuclear construction. Parliament passed the revised Energy Act in 2016. On May 21, 2017, with Leuthard serving as Federal President, Swiss voters approved the strategy by 58.2 percent.

They were not reckless, and they were not naive about speed. Only six months earlier, on November 27, 2016, the same electorate had rejected a Green Party initiative that would have forced a much faster shutdown, by 2029. The 2017 yes was the moderate Swiss path: keep the plants that were already running for as long as the nuclear regulator, ENSI, judged them safe; refuse to build new ones; cover the gap with Alpine hydro, rooftop solar, efficiency, and the European market. In a country that already generated most of its electricity from hydro and nuclear, and that prided itself on climate diplomacy, the package looked like prudence rather than rupture. Germany was exiting nuclear on a similar timetable. France would always have reactors to spare. The Alps stored water. For a consensus democracy working from 2011’s information, the strategy was a rational answer to a real fear.

Segment 3 — The Implementation

The revised Energy Act entered into force on January 1, 2018. It did not padlock reactor doors. It forbade new builds and let existing units run on a safety-based lifetime, which is why Gösgen, Leibstadt, and the two Beznau units are still on the grid years later. The first physical subtraction was Mühleberg. BKW had already decided, in 2013, to close it at the end of 2019 for a mix of economics, politics, and upgrade costs; the ceremony on the Aare was the moment the strategy became megawatts. Proponents in the Social Democratic Party, the Greens, and parts of the center argued that efficiency and a renewable build-out would more than replace what retired, and that Switzerland would remain an island of clean power in the heart of Europe. Official scenarios from the Swiss Federal Office of Energy sketched a 2050 system dominated by hydro and new renewables, with imports as a residual, not a foundation.

Skeptics were specific, and they were not only the plants’ home cantons. The Association of Swiss Electricity Companies, along with Axpo, Alpiq, and parts of the business federation economiesuisse, warned about a winter electricity gap — the Stromlücke. Swiss demand peaks in the cold months, when heat pumps and resistance heating run, solar output collapses, and Alpine reservoirs are being drawn down rather than filled. Nuclear had been the quiet winter partner of hydro: firm, domestic, nearly carbon-free. The strategy assumed that “the European market” would supply whatever winter energy Swiss generators could not. It did not pair the nuclear ban with a committed build of replacement firm low-carbon capacity — no new reservoirs on the scale required, no domestic gas with capture, no long-duration storage at fleet scale, and, by design, no new reactors. Early years looked calm. Hydro had good years. France was still a reliable exporter. Solar installations began to climb. The gap was a graph in a winter outlook, not yet a political emergency.

Segment 4 — The Unintended Consequences

Physics does not grade annual press releases. Switzerland consumes on the order of 55 to 60 terawatt-hours of electricity a year, with nuclear historically supplying roughly a third of generation and hydro most of the rest. Mühleberg’s closure removed about three terawatt-hours of winter-capable output. That is not a large number on a European scale, and it was never supposed to be replaced by a coal plant in Aargau. It was supposed to dissolve into a greener market. The market, in January, after dark, does not work that way. When Swiss reservoirs are falling and rooftop solar is producing near nothing, the next megawatt Switzerland can buy is whatever is on the margin in the coupled grids of Germany, France, Austria, and Italy. In many winter evening hours of the late 2010s and early 2020s, that marginal plant was not a French reactor or a German wind farm. It was hard coal or lignite, sometimes gas.

The causal chain is almost mechanical. Close or withhold firm low-carbon supply in a winter-peaking alpine system. Leave the residual to cross-border trade. Watch Germany, Switzerland’s largest electrical neighbor, run down its own nuclear fleet while lignite and hard coal remained available to set the price and the carbon content of the last kilowatt-hour. Switzerland’s production mix stayed among the cleanest in Europe; its consumption mix did not, not in the hours that mattered. Analyses of cross-border flows and of the residual supply mix — the kind of work grid planners and independent carbon-intensity trackers rely on — showed the carbon intensity of Swiss electricity consumption rising during winter peaks as nuclear output fell and imports rose. Electrons at the Laufenburg and Bassecourt interconnectors do not carry passports, but the fuel that produced them does. A country that could publish almost spotless generation statistics was, on cold nights, partly powered by the Rhineland and the eastern German lignite districts.

Second-order effects stacked on the first. Because international climate inventories are territorial, the extra tonnes were Germany’s problem on paper and Switzerland’s only as an import footnote. Guarantees of origin let suppliers market “Swiss hydro” even when the physical residual mix was dirtier, so households could feel they had bought clean power while the grid was balancing with coal. The long-stalled electricity agreement with the European Union left Switzerland less fully coupled than its neighbors, which complicated both security and the clean sharing of surplus. Inside the country, the missing winter energy became a planning assumption: more imports, more political anxiety, and eventually more fossil backup. Third-order effects arrived with the weather and with other people’s accidents. Dry years cut hydro. In 2022, after Russia’s invasion of Ukraine, gas prices exploded just as a large fraction of the French nuclear fleet was offline for corrosion repairs. Germany, heading toward the April 15, 2023 closure of its last three reactors, burned more coal to keep its own system upright. Switzerland, which had treated the European market as a clean cushion, discovered it was a cushion stuffed with whatever fuel the continent still had.

The human texture of that winter was not abstract. In Aarau, Swissgrid’s operators watched evening ramps that used to be absorbed by Mühleberg and the French interconnector. In the Bernese countryside, the Mühleberg site went from employer and taxpayer to a decommissioning project scheduled to last into the 2030s. In German mining regions, lignite plants that climate policy had slated for a managed decline found themselves, once again, the machines that kept a rich neighbor’s heat pumps running after sunset. Swiss officials spoke carefully of “supply security.” Opposition politicians and utility engineers used a blunter word they had been using since 2011: gap. The irony was not that Swiss voters had wanted more coal. They had wanted less nuclear risk and less carbon. The design treated annual renewable energy as a substitute for firm winter capacity, and treated a border as the place where carbon stopped counting.

Segment 5 — The Aftermath

By late 2022 the Federal Council was no longer talking only about 2050. It ordered a hydropower reserve, paying operators to hold water back for the darkest weeks, and it contracted emergency reserve generation — gas turbines at Birr in Aargau, built as insurance against a shortage that the 2017 strategy had not supposed to exist. The reserve was expensive for the hours it might run, and it added a further twist: a nuclear phase-out justified in part on climate grounds had summoned fossil backup onto Swiss soil. Solar, to the country’s credit, accelerated sharply after 2022, especially on rooftops and alpine installations, and summer afternoons now produce genuine surplus. Winter evenings did not change their minds.

Public opinion and parliamentary arithmetic moved. Albert Rösti, who took over the energy department in 2023, opened a political path toward lifting the ban on new nuclear construction, while operators of Beznau, Gösgen, and Leibstadt planned for long-term operation under ENSI’s safety regime. By 2024 and 2025 Bern was formally revisiting the central legal pillar of Energy Strategy 2050. That reconsideration is itself unfinished business, not a fairy-tale reversal: new reactors, if they come at all, take more than a decade, and the plants still running cannot be wished into being twice. The current state is a country with a cleaner summer, a still-tight winter, a nuclear fleet in overtime, a gas reserve it did not want, and a more honest conversation about the difference between generating clean electricity and consuming it.

Segment 6 — The Lesson

Three principles travel well beyond the Aare. First, annual energy targets are not the same product as a kilowatt-hour at 7 p.m. in January; if you retire firm low-carbon capacity before firm low-carbon replacement exists, the market will buy whatever still burns. Second, emissions follow the marginal plant, not the territorial inventory — a border is not a climate policy, and production-based accounting will always invite leakage that looks like virtue. Third, incentive structures and certificate systems will decouple what people think they bought from what the grid actually used, unless physics and bookkeeping are forced to tell the same story. The Swiss were not foolish. They were sequential. As governments write 2030 and 2035 decarbonization targets, and as data centers and heat pumps add new winter load from Dublin to Virginia, the question Mühleberg left on the table is simple: when the clean plant that works in the dark closes, who, exactly, do you think will be on the margin?

💬 Reply to this email — Patrick reads every one.

Share: X · LinkedIn · WhatsApp

Forwarded this email? Subscribe here — it's free.