It's a matter of conjecture whether declining world coke production is the cause or the effect of the growing practice of replacing coke in blast furnaces with injected pulverized steam coal.

But what is undeniable is that U.S. coke production has been steadily declining since the 1950s. In the last 15 years alone, it is estimated that domestic coke production dropped more than 50 percent in 1994, to 22 million tons from 46 million tons in 1980.At the same time, the injection of pulverized coal, granular coal and natural gas into blast furnaces has increased dramatically as injection technology becomes refined.

"A lot of companies are shutting down their coke plants because of increased environmental regulations," said Rick Shulkosky, manager- manufacturing and technology with the American Iron and Steel Institute based in Washington.

Under the U.S. Clean Air Act of 1990, coke producers were given their choice of two separate options for cleaning up their operations. The first option permitted coke plants to meet "maximum achievable control technology" standards by the end of 1993 and "lowest achievable emission rates" by the end of 1998. In return, they were allowed to postpone complying with the EPA's stringent "residual risk standard" until the year 2020.

The second option allowed coke plants to meet "maximum achievable control technology" standards by the end of this year but, in turn, they must comply with the "residual risk standard" by the year 2003.

At least partially as a result of an uncertain future, coke supply together with rising coke prices, steelmakers throughout the world are scrambling to reduce their reliance upon the commodity.

And in the last decade, there has been a resurgence of interest in a technique known as pulverized-coal-injection, which reduces blast furnace coke requirements on an almost pound- for-pound basis although the use of coke cannot be totally eliminated.

Using coal injection, a wide variety of coal types can be employed, including steam coal with varying degrees of sulfur content. The injection process involves crushing, grinding, drying and sizing the coal particles before they are stored in a pressurized tank. The crushed coal is then pneumatically pumped to injection units that meter the proper amount of coal into the blast furnace.

Babcock & Wilcox and Armco Steel together pioneered the world's first commercial application of pulverized coal injection at an Armco Steel Ashland, Ky., steel mill in 1963. But despite this early lead, pulverized coal injection has been slow in gaining acceptance in the United States although its use is common in Europe and Asia.

There are currently more than 70 pulverized coal injection systems in operation outside the United States and it is reported that 25 of the 32 blast furnaces in Japan are using coal injection. But that global imbalance is rapidly changing with virtually all major U.S. steelmakers investing in the technology.

Inland Steel and U.S. Steel installed pulverized coal injection systems at several of their mills a few years ago.

Bethlehem Steel Corp. has invested an estimated $192 million to install coal injection equipment at its Burns Harbor, Ind., plant. Of this amount, the U.S. Department of Energy (DOE) has contributed $30 million toward the project.

The DOE made the investment in an effort to determine if granular coal, which is approximately the fineness of granulated sugar will work as well as pulverized coal that is crushed to a talcum powder consistency. Preparing coal to a granular form rather than a powder will offer significant savings to steelmakers. Bethlehem Steel's new installation will be able to use both granular and pulverized coal.

Ian Cox is director of project management for ATSI Inc. of Amherst, N.Y., the company doing the engineering work for Bethlehem's Burns Harbor plant. Mr. Cox said the primary reason pulverized coal injection has never been widely adopted in the United States was because of the historically low cost of coke.

"Compared to using coke, injecting coal 30 years ago was considerably more expansive. But today we're developing a granular coal system that requires much less preparation than pulverized coal," said Mr. Cox.

He added that there are no qualitative differences in steel made from coke or from injected coal.

"If you're in control of your process, there's no reason for there to be any difference in quality between using coke or coal," said Mr. Cox.

The impact of widespread coal injection on the U.S. steam coal market is expected to be minor. Coking coal has never comprised more than a few percent of total U.S. coal production.

The National Coal Association (NCA) based in Washington estimates 1995 U.S. coking coal consumption at 32 million tons, compared with 838 million tons of steam coal used by U.S. electric utilities. The NCA expects coking coal exports to account for an additional 46 million tons this year.

Thus, even in the extremely unlikely event that coal injection becomes adopted worldwide in the next decade, it should have a minimal impact on the supply or pricing of steam coal.