TWO MEN WERE KILLED a few days ago when the roof collapsed in the Scotia Coal Company's mine near Oven Fork, Ky. It's the same mine in which 26 men were killed three years ago. In this week's accident, a their miner was apparently saved by the protective canopy of the machine he was operating.The mining company has been contesting the rule requiting canopies.

A day earlier, and light years distant, the federal Nuclear Regulatory Commission had ordered five power reactors shut down. a design error had come to light, suggesting that some of the piping might not withstand a severe earthquake. How severe? Reactors are typically designed to survive the most powerful earthquake that might occur over a span of 10,000 years. The NRC is doubtless correct in thinking that the American public is intolerant of risk in reactors.

The cases of the coal mine and the reactors are tied together, of course, by the demand for electrictity. Coal generates almost half of the nation's electric power, and nuclear reactors about one-eithth. In all matters concerning coal, this country is comparatively relaxed about health and safety protection. But in everything that touches nuclear power, it is increasingly cautious. Americans are highly selective in their perceptions of risk.

Perhaps part of the explanation is that nuclear power is new, carrying with it unfamiliar and especially frightening new dangers and evoking new standards. Coal has been around for a long time and, though its costs to human life are large, they are familiar. Coal became part of our lives when people were poorer and life was cheaper. The usages established then still seem to influence attitudes today. It surprises no one that frim time to time a mine should collapse and men die. Nor does it surprise anyone that, after ugly accidents, mines should reopen and men continue to work in them.

But, in terms of health hazards, mining coal is less dangerous than burning it. Statisticians have repeatedly shown that the air pollution resulting from coalfired power plants results in shortened life spans in the populations affected by it.

A nuclear reactor makes a small but measurable contribution to the radiation to which the public is exposed -- another form of dangerous pollution. How does this hazard compare with that of coal? There have been many studies and, while the numbers are not precise, they demonstrate clearly that coal imposes a much greater cost to human health, perhaps 100 times as high -- measured in terms of premature deaths.

A lot of Americans are deeply uneasy about reactors, wondering whether one will explode disastrously or perhaps begin silently leaking readiation. Yet the greater hazards of coal are generally met with the thought that it's one of the things that you just have to live with.

The sudden shutdown of the five reactors reflets the financial caution that now characterizes nuclear regulation. Those reactors have been turned off at substantial inconvenience to the antion, in the midst of an oil shortage and at heavy costs to the customers. Not many people would want it otherwise. But it's curious to think that, if the same health standards were applied to coal and the power plants that burn it, half of the country would be dining by candlelight tonight -- and not voluntarily