COKE, A FUEL USED mainly in iron and steel manufacture, is essentially pure carbon. If it came out of the ground that way, life would be a lot simpler—and cleaner. Unfortunately, coke starts out as bituminous coal, a mixture of carbon and various impurities. Separating the impurities is not particularly hard; you just heat the coal to drive them off. The problem lies in disposing of the unwanted portions. For decades each new solution simply replaced an old environmental peril with a new, unexpected one.

The earliest apparatus used in coke manufacture was the beehive oven, which originated in western Pennsylvania in the 1850s. Each oven was a circular, domed chamber in which coal was baked for about forty-eight hours, using the heat provided by burning volatile by-products as they escaped. Smoke and unburned by-products were vented into the atmosphere through an opening at the oven’s top, leaving a layer of coal dust, ash, and particles on nearby woods and fields. One botanist wrote of “the general wretchedness of everything of the nature of a shrub or tree” in coke-producing areas. Cities and towns banned the ovens or required very tall stacks, and courts often awarded damages to farmers.

Early in this century by-product ovens started replacing beehives. The by-product oven, a narrow retort oven constructed in batteries, was faster and more efficient and allowed the use of lower-grade coal. It appeared to solve the emissions problem as well: Volatile elements were captured by pipes at the top of the oven, condensed, separated, and sold. By-product oven construction accelerated after 1915, when U.S. Steel discovered that gas and tar from coke ovens could produce important fuel efficiencies in its integrated steel mills. A war-induced cutoff of organic chemicals from Germany also helped.

For ease of transportation and waste disposal, most byproduct plants were located on the banks of rivers or lakes, usually close to mills and to metropolitan areas where the by-products could be sold. However, the liquid waste stream always contained large concentrations of ammonia, cyanide, and phenolics, which produced unpleasant tastes and odors in drinking water. Beehive ovens had spewed out much more waste, but they had spread it over much larger areas that were sparsely populated. By-product ovens, by concentrating their disposal in urban rivers, made their reduced waste streams especially dangerous and annoying.

If the water was chlorinated, things got even worse. Phenol wastes interacted with the chlorine to create an assortment of extremely foul-smelling compounds, making the water undrinkable. During a 1925 typhoid outbreak in Ironton, Ohio, for example, offensive tastes in the municipal water supply drove residents to drink untreated water. Thus two technologies, each regarded as an environmental improvement, had combined to produce a new health problem.

Municipalities and health departments hurried to seek relief. In 1918 McKeesport, Pennsylvania, secured a court injunction against U.S. Steel’s new Clairton Coke Works, the world’s largest such operation, ordering the plant to stop polluting the city’s water supply. In the mid-1920s Ohio and Pennsylvania passed regulations limiting the discharge of industrial wastes into rivers used as public water supplies, and all the Ohio Valley states signed an agreement seeking the same end. State health departments reached agreements with coke firms, and by 1929 seventeen of nineteen valley by-product producers had installed phenol control devices.

Most of these plants adopted a technique that used waste water to quench the glowing coke, thereby vaporizing the wastes. Engineers thought that quenching would destroy the pollutants, but they soon learned that the volatile components were instead being steam-distilled and discharged into the atmosphere. The problem that had plagued beehive ovens was back. Coke manufacturers, having invested large sums in quenching equipment, said there was no other way to meet both environmental and economic constraints. With only a very limited understanding of the effects of airborne contaminants, the air became a preferable sink to the water, where the effects of pollution were apparent every time someone turned on the tap. It would take another half-century for regulators to finally challenge, and ultimately change, industry practice, by requiring pretreatment of quenching water to remove wastes.