The Washington area's largest home insulation firm said yesterday it is abandoning production of a controversial brand of insulation and replacing it with one that has sharply lower insulating values.

The decision, made by officials of Davenport Insulation Inc., a subsidiary of the Washington Gas Light Co., leaves unanswered questions that government and industry officials has raised about the effectiveness of the old product, marketed under the name of "Thermlo-K."

However, the new product, to be known as "Thermlo-K II," will be advertised with insulating values that scientists have said is more in line with other insulating products that also are made from finely ground newsprint.

In an interview yesterday, Carol V. Davenport, president and founder of the insulation company, said his decision to stop making the old product had nothing to do with the controversy. Rather, he said, it was based on the tightening supply of boric acid and borax, two of the chemicals that are added to newsprint to make cellulose insulation.

Faced with a shortage of those chemicals, Davenport said, his firm could not continue to produce Thermlo-K at its plants at Lorton and at recently opened plants in Albany, N.Y., and Valdosta, Ga. Davenport's company spent $100,000 to develop the new product, which he said is "completely different chemically" from the old product.

Both the old and new Thermlo-K are sprayed into homes with a nozzle attached to a garden hose, a technique that has prompted some of the controversy over the product.

Many scientists interviewed by The Washington Post in the fall questioned Davenport's claims for the product's insulating values - measured in units called "R" values. The scientists said that spraying a product with water would be more likely to reduce its R value than to increase them, as Davenport claimed.In advertising, Davenport listed Thermlo-K's R value as 5.2 per inch - higher than any other cellulose insulation in the country - and a level that some scientists said was impossible under the laws of physics.

Davenport refused yesterday to back off those claims, but he said the company never had the product tested as he had first said in September and as a Washington Gas Light spokesman had promised again in November. Instead, he said the new product has been tested by the National Research Council of Canada. Davenport's planned ads will describe the council as "one of the world's most prestigious laboratories."

The tests gave the new product an R value of 3.5 per inch when sprayed on ceilings and 3.7 per inch on walls, Davenport said. Those levels match test results for many other cellulose insulation products.

Davenport had agreed last fall, under pressure from the National Cellulose Insulation Manufacturers' Association, a trade group, and other building industry groups, to have the original product tested by an independent Massachusetts laboratory. He backed off those tests when a Washington Post story quoted some of the laboratory's officials as questioning the Davenport claims.

Later, a spokesman for Washington Gas Light said Davenport had decided to move the tests elsewhere because he wanted a lab "he would feel more comfortable with." Davenport said yesterday that he dropped plans for the tests altogether "because we were out of production with that product."