THE ACCIDENT at the Three Mile Island nuclear reactor is an event that, by every prior calculation, should have been exceedingly improbable. Some of the questions are pretty obvious: Are those safety calculations reliable? Do they require reconsideration? Was it mechnical failure, or human error? If an operator made a mistake, it is not going to be corrected by redesigning pump and pipes.

The reactor has vented some rarioactive gas into the atmosphere of central Pennyslvania, and yesterday it was still emitting radiation through its concrete shield walls. In boths cases, the levels of radiation were low. Federal officials said that they were of the same magnitude as background radiation from natural sources. The exposure to people outside the plant was apparently similar to the exposure that an airline passenger might receive in a flight at high altitude. But, as the Department of Health, Education and Welfare warned earlier this month, current research indicates that any increase in radiation is harmful, and low-level radiation from natural sources may be an important cause of cancer. The health danger to the general to be small. But it is not insignificant.

There is no way, unfortunately, to generate electricity without risk. Good engineering and wise regulation can reduce that risk, but never to zero. Hydroelectric plants are generally the safest of all the possibilities, but even dams have been known to burst. Solar technology is available for heating bath water. But generating power from sunshine, on a commercial scale, is decades in the future. Nearly one half of the nation's electricity comes from coal-fired generators, and the process of burning coal inevitably puts highly toxic gases into the air that we breathe. A coal furnace represents a substantially greater threat to public health than a uranium reactor of the same size. It would be sadly irnoic if the only effect of the accident at Three Mile island were to shift American utilities away from uranium and more heavily than ever onto coal.

Since there are no risk-free solution, the most sensible course is to continue to rely on a variety of different sources of power, including nuclear-and to keep developing them slowly and cautiously. But if the development of nuclear power is to proceed slowly and cautiously, two things must happen. One is that the need for power must be limited, since the faster Americans increase their demand for addnitional electricity, the harder it will be to remain prudent. The other thing that must happen if nuclear power is to be developed at all is that the public alarm verging on hysteria concerning the dangers be taken account of and honestly answered. That is why the explanations of the Three Mile Island accident will be to be exceedingly thorough, persuasive and public.