Since Fukushima, Japan’s nuclear meltdowns in 2011, the world has taken a turn away from atomic energy. Like Chernobyl (which triggered a nuclear phase-out in Italy) and Three Mile Island (which caused Sweden to abandon nuclear energy) before it, Japan’s triple-meltdown prompted a number of countries to reconsider the power source. German Chancellor Angela Merkel reversed her pro-nuclear stance, and shut down around half of the country’s reactors, vowing to phase out the energy source entirely within 10 years of the accident. Switzerland followed suit, opting for a more gradual phase-out. Taiwan ended up pressing ahead with nuclear projects, but not after a contentious debate.

However, while nuclear power is controversial and can have disastrous local effects if it goes wrong, an abrupt withdrawal from the carbon-free energy source is guaranteed to hike emissions in the near term.

Late last month, the U.S. Energy Information Administration published an analysis of the long-term effect on carbon emissions of phasing out two of the most controversial sources of energy: coal and nuclear power. The report shows that cutting down on coal use can more than compensate for the impact of ditching nuclear power.

The EIA’s analyses are generally quite conservative, and their so-called “reference,” or base case, may be far off the mark given the speculative nature of predicting energy use and carbon emissions decades into the future. But what they found was that if regulators limit nuclear reactors to 60 years of operation (currently, “operators are preparing applications for license renewals that would allow operation beyond 60 years,” the administration wrote), total carbon dioxide emissions in the U.S. from electricity generation would be 4 percent higher than if the nuclear plants are kept in operation as long as possible. (For context, about 70 percent of all electricity generated in the U.S. comes from burning fossil fuels, while electricity generation contributes to around 40 percent of total emissions.)

(Credit: U.S. Energy Information Administration)

To compensate for reactor closures, the EIA predicted that natural gas-fired generators would produce 13 percent more power (natural gas was seen as the most likely alternative to nuclear “in all cases examined because of competitive fuel prices and relatively moderate capital costs”), while renewable generation would increase by 5 percent.

As an example of what happens when nuclear reactors are abruptly taken offline, Vox explained the outcome of two reactors in California unexpectedly shutting down:

Back in February 2012, Southern California Edison shut off two nuclear reactors at the San Onofre plant after finding cracks in the steam generator system. (A year later, the company announced that it would retire the reactors for good, deciding the repair and licensing process would take too long and involve too many lawsuits.)

That plant was massive, providing about 8 percent of California’s electricity. So the state went on a frenzy of construction, building mostly new natural gas units and some wind units. In the end, however, fossil fuels were the easiest to deploy. Overall carbon-dioxide emissions in the region rose by 9.2 million tons in the following year — equivalent to putting an extra two million cars on the road.

But while accelerated nuclear retirement will increase emissions by 4 percent, accelerated coal retirement has the potential to more than compensate for the reduction. The EIA found that carbon emissions would drop a whopping 20 percent if the costs of mining and transportation rise (by 68 percent by 2040) as well as the costs of actually burning the fuel (3 percent a year are assumed, as opposed to the current flat costs) — although these costs could easily be simulated through the right kind of regulation and taxation.

As with advanced nuclear retirements, much of the slack would be made up for through natural gas-fired generators. But since natural gas is a much cleaner-burning fuel, the 19 percent increase in gas-generated electricity combined with a 10 percent increase in renewables would still yield a one-fifth reduction in total electricity-related carbon emissions.

So while the best-case scenario for carbon emissions — which are a very different thing than the highly localized impact of a rare nuclear meltdown — would be to keep atomic energy humming while hiking the cost of coal to force many plants into an early retirement, ultimately the effect of cutting coal generation is far more important to overall emissions.