Here’s some bleak news for the coal industry: As much as 65 percent of the U.S. coal fleet could find itself under threat in the years ahead, thanks to cheap natural gas and stricter air-pollution regulations.

That’s according to a new peer-reviewed study by three researchers at Duke’s Nicholas School of the Environment, who take a detailed look at the costs of operating both coal-fired power plants and natural-gas plants around the United States.

Their conclusion? Coal power is far more economically vulnerable than most analysts have realized to date. Here’s why:

Cheap natural gas is crowding out coal: Already, a glut of cheap natural gas from shale deposits in Texas, Ohio, Pennsylvania, and elsewhere is upending the electricity sector. The researchers found that around 9 percent of the U.S. coal fleet has become uneconomical — it’s now cheaper to burn natural gas for electricity than to keep running those coal plants, which are now slated for retirement.

The chart below sums up the ongoing shift quite vividly. Many coal plants are now operating at far lower capacity in 2012 than they were back in 2007:

And new pollution rules could accelerate that shift: But the shift to gas could become even more dramatic in the years ahead. Under the Obama administration, the EPA has been tightening various standards for air pollutants, including sulfur dioxide, particulate matter, and mercury. By 2016, power-plant operators will have to start installing costly new emissions controls at their coal units to comply with these rules. And that will further sway the economics of power generation.

By looking carefully at operating costs, the Duke researchers found that an additional 56 percent of the U.S. coal fleet could become more expensive than natural gas, assuming the ratio of gas prices to coal stayed around its current level. That poses a huge challenge to the coal industry — far bigger than previous analyses have suggested.

“That was the most surprising thing to come out of this analysis,” says Duke’s Lincoln Pratson, who led the study. “There has been a lot of focus on those 9 percent of coal plants set for retirement — mostly smaller, older plants. But what were seeing is that there are a very large number of younger plants that would become more costly to run than the median natural-gas plant if they were to upgrade their emissions controls.”

That said, there are some caveats: That doesn’t mean 65 percent of the coal fleet is set to shut down in the years ahead. This analysis all depends on a number of assumptions. First and most obviously, natural gas prices could start rising again — the early glut of gas from the fracking boom already appears to be leveling off. That would help many coal plants hang on.

Yet, surprisingly, Pratson and his colleagues found that natural gas would stay competitive even if the ratio of natural gas to coal prices rose from its current level of 1.7 to 4. ”Even in that case, we wouldn’t see the situation we had [in the early 2000s] where every coal plant was cheaper to run than every single natural gas plant,” Pratson says.

The study makes a few other debatable assumptions as well. For one, it assumes that power plant operators will comply with the EPA rules by installing new emissions controls — rather than, say, buying allowances under cap-and-trade systems like the acid-rain program. On the flip side, the study doesn’t take into account further potential EPA regulations on cooling water or coal ash. As such, says Pratson, the analysis is offering a “mid-range estimate” on coal’s vulnerability.

As a side note, the study also helps referee a contentious political debate. During the 2012 campaign, there were two big theories for what, exactly, was killing the U.S. coal industry. Many conservatives blamed the EPA’s air pollution rules, part of President Obama’s “war on coal.” Other analysts largely chalked it up to cheap natural gas — this was just the market at work.

This new study suggests that both are crucial factors, and tries to look at how, precisely, natural gas and the EPA will interact with each other in the years ahead.

Further reading:

–The decline of coal, in three charts.

–The shift from coal to natural gas has helped drive down U.S. carbon-dioxide emissions. But natural gas will only be effective in tackling climate change if the industry gets a handle on methane leaks.

–What happens to America’s coal if we don’t burn it? Mining companies will try to ship it abroad, mainly to China. But that all depends on whether they can build new export facilities in the Pacific Northwest.