The number of U.S. nuclear power plant workers exposed to maximum radiation doses, and the accompanying risk of cancer, soared last year as plants underwent more repairs, alterations and maintenance than usual, an environmental group reported yesterday.

The Environmental Policy Institute, a nonprofit Washington-based research group, released a May 28 memorandum of the Nuclear Regulatory Commission radiation control section which said that the 68 operating atomic power plants exposed their work forces to 35 percent more radiation last year than in 1979, even though there was only one new plant.

A record number of 80,200 workers were exposed to radiation last year, compared with 64,073 in 1979. Average exposure was an all-time high of 791 person-rems per reactor, or 53,797 person-rems total, compared with 39,759 person-rems the year before.

Person-rems measure the sum of all exposure to all workers, in the same way that work-years measure the total time put in by all workers.

The exposure rate has nearly quadrupled since 1969, when it was less than 200 person-rems per reactor.

The Surry nuclear plants of the Virginia Electric and Power Co. had the second-highest rate of worker exposure among pressurized water reactors, emitting 1,900 person-rems, more than double the national average. The San Onofre (Calif.) plant had the highest rate.

The policy group noted that the amount of radiation each worker receives has declined substantially since 1969, from 1.03 rems to 0.73 rems per year, well within the NRC's limit of 5 rems.

But the number of workers exposed to radiation has risen dramatically, from 145 at an average plant in 1969 to 1,010 last year, the report stated.

This is because "nuclear plants have each 'used up' more and more radiation workers" as more get exposed to the maximum radiation regulations allowed early in the year and have to be replaced by other, often temporary, workers.

The institute said the findings are "extremely serious" in terms of the number of people exposed to radiation, given current research on the effects of low-level doses. If the current level continues for 20 years it means three extra genetic defects per 100 births, the study said.

The findings also rebut yearly explanations from industry that "some particular problem in the reactors has been given a 'one-shot' fix and won't happen again," Fred Millar of the institute said. "The exposure comes from increased radioactivity in permanent nuclear plant components."

William E. Kreger, the NRC's assistant director for radiation protection to whom the memo was directed, confirmed that assertion in part.

"A slowly increasing amount of radiation inside the primary reactor cooling system was noted in the mid 1970s," he said. Research traced it to corrosion materials, and new kinds of water chemistry are beginning to deal with the problem, he said.

"We now think it will not continue to build at this rate" and that 1980 and 1981 may be the peak exposure years, Kreger continued.

Dave Harward, environmental projects manager at the Atomic Industrial Forum, the industry trade association, blamed the high radiation output on changes required after the accident at Three Mile Island.

Some of the rise from 1969, he said, results from aging plants. "When a plant is new it doesn't require as much maintenance," he said.

Millar rejected the Three Mile Island argument. "Most of what's been done so far has been simply paperwork. Installation of the hardware hasn't really got under way in most places," he said.

Kreger said the Surry plant and some of the other "hotter" units have been repairing and replacing corroded steam generators in which there has been "a fair amount of radioactivity buildup." However, he added, high radiation output was "a source of concern" to the NRC.