The Environmental Protection Agency is considering relaxing its standards for the most controversial of the major air pollutants, sulfur dioxide.

A product of burning coal and oil, sulfur dioxide combines with water vapor at high altitudes to cause acid rain, which kills fish and plants. The EPA is concerned about acid rain and wants to tighten control over sulfur dioxide emissons, a tough enough job in the face of growing national reliance on using coal for energy.

But according to an internal memo obtained by the Washington Post, a major decision is at hand that could end up loosening the standards instead.

The state of Ohio wants permission to average the sulfur dioxide emissions of its utilities and industries over 30 days to determine whether they violate the rules instead of averging them over 24 hours as is required now.

According to the Dec. 6 memo from Walter Barber, EPA's director of air quality planning and standards, to EPA Deputy Administrator Barbara Blum, such a change would be "a significant relaxation in emission limits" and would put Ohio way over EPA's air quality standards. It cannot be allowed Barber wrote.

But to hold Ohio to the 24-hours standard won't work either because under the current EPA system it is unfair, the memo says. "The regulation is set assuming that emissions from the plant are constant and are always at the maximum level," Barber wrote. "The assumption is wrong and increasingly it is becoming a source of confusion and conflict."

It seems that nobody at EPA knew enough about coal 10 years ago when the regulations were written for the Clean Air Act.

Everyone assumed then that coal from one mine or region was more or less uniform in sulfur content and that sulfur dioxide emissions from burning it would be uniform. The regulations, therefore, said nothing about the period of time that emissions had to be watched. It was assumed that one quick smokestack sample would be typical whenever it was taken, and sampling was the method chosen to determine compliance with the rules.

Now, however, EPA and everyone else understands that the sulfur content of coal not only varies widely from mine to mine but also from lump to lump within a single mine. One 24-hour burn can produce a high average sulfur dioxide level while the next day's average may be very low.

Environmentalists were pleased by this discovery, arguing that the chance of overrunning the limits forced utilities to use generally lower-sulfur coal so that any one day's burn wouldn't be too high.

In practice, however, utilities and industries use the coal that's available and make sure, according to EPA officials, that some low-sulfur stuff is burning on the day the EPA watchdogs arrive for the stack sample.

"Stack tests don't work very well," said Barber in an interview. "In practice we don't do many of them."

Instead, EPA has been relying on reports the utilities file on their own emissions, and these are usually calculated on a long-term average since the Clean Air Act is silent on the averaging period. (Ohio is an exception with its 24-hour requirement since its compliance plan was forced on it by EPA last year after the state failed to come up with its own.)

As a result, Barber's memo said, current national practice is putting 85,000 to 95,000 tons of sulfur dioxide into the air every year. Requiring everyone to adopt the 24-hour period and somehow enforcing it would cut emissions to 60,000 tons.

EPA would like to do that, but it can't because the 24-hour period is unfair under current practice, the memo added.

But EPA would like very much to keep Ohio happy, partly because the state is a political battleground in an election year and partly because major legal action would likely follow any decision that forced the state to go outside its boundries for coal.

So EPA has come up with a new way to figure sulfur dioxide emissions that would be realistic about coal and keep a 24-hour standard in Ohio while allowing a 30-day or even a yearly standard elsewhere. The problem is that the net effect would be to legitimize the status quo, in effect eliminating the possibility that sulfur dioxide emissions can ever be reduced below current levels.

The new method is called the "expected exceedances" or ExEx method. It involves a sliding mathematical scale of sulfur content in coal combined with computer calculations on how often a plant with that coal is likely to exceed air quality standards at any given point on the ground nearby. l

Barber's memo recommends denying Ohio's application for a 30-day averaging period while announcing that the ExEx method is to be installed nationwide. "However, it is not appropriate that this change be interpreted as a lack of concern by EPA over sulfur dioxide," the memo continues. "I recommend that we . . . offset it by enforcement initiatives and strengthening of modeling guidelines."

EPA Administrator Douglas Costle is expected to try to resolve the conflict within the next two weeks.