Union Carbide Corp.'s safety chief, responding to critical federal reports, today said that its pesticide plant at nearby Institute is "completely safe" and that comparison of it to the company's plant in India is "irrelevant."

Jackson B. Browning, vice president for health, safety and environmental affairs, said a different system was used to cool methyl isocyanate at the plant in Bhopal, India, where a massive leak of the highly toxic chemical last month killed at least 2,000 people.

Although the Bhopal plant was modeled after the one here in Kanawha Valley, Browning said freon was used there to cool the chemical, while water is used at the plant near here.

Because of that difference, he said, a warning from a Union Carbide safety team about a possible runaway reaction in the cooling procedure at Institute, reported nearly three months before the tragedy in India Dec. 3, was not passed to plant officials in Bhopal.

Browning would not respond to a report by an Indian scientist earlier this month that he believed that water leaking into the methyl isocyanate tanks there caused a runaway reaction that turned the liquid chemical into deadly clouds of poison gas.

Speaking at a news conference in Danbury, Conn., Union Carbide's corporate headquarters, Browning said the "serious allegations" raised Thursday by Rep. Henry A. Waxman (D-Calif.) about the possibility of a "runaway reaction" at the Institute plant were removed by "a simple change in procedure."

Until the last 3,000 pounds of methyl isocyanate was used Thursday, the Institute plant was the only place in the United States where the chemical was manufactured, although it was shipped from here to several locations. It is a key ingredient in several pesticides.

Browning also disagreed with a report Wednesday by the Environmental Protection Agency that 28 instances of methyl isocyanate leaks had occurred at Institute in the last five years.

Browning said there have been no leaks at the plant sizable enough to require reporting to the agency. He said a report that 840 pounds of methyl isocyanate was discharged at the Institute plant Dec. 31, 1982, was incorrect.

EPA spokesman Dave Cohen said the agency's findings were taken last month from Union Carbide records.

"Those figures were not made up," Cohen said. "They didn't come out of the air."

Cohen said the EPA is "in the middle" of a followup investigation. "The first time around, the data that we were given led to that conclusion," he said.

Carl Beard, director of the West Virginia Air Pollution Control Commission, said H.J. (Hank) Karawan, manager of the Institute plant, called him this afternoon and told him of Browning's statement that the December 1982 discharge report was incorrect.

"Carbide takes the position that the leak was actually five pounds," said Beard, who said he is keeping "an open mind" on what might be the correct figure. "There is some confusion," he said, adding that the company's statement "doesn't mean anything. We want the EPA to review its findings."

Beard said that, because his agency must rely on information provided by industries it regulates, "you need some skepticism. If not, you are going to be burned."

Browning said the mistake in calculation "was our fault." He said Union Carbide officials misidentified the tank in which the leak occurred.

The EPA was told that a line broke leading from a tank that contained 14,000 pounds of a mixture that was 6 percent methyl isocyanate, although the actual percentage of the chemical in the tank was "minuscule," he said.