Agglomerative aspects of metallurgical coke production in the United States: Facility grouping, combinations, and spatial structure

The primary intent of this study is to examine the spatial patterns and spatial morphology of the United States metallurgical coke industry in an attempt to determine if the coke industry and other industries with trading linkages with it, in addition to the iron and steel industry, spatially group and combine together in such a way as to form a class of industrial agglomerative sub-structures and sub-groups centered around the metallurgical and the hydrocarbon (hydro-met) processing technologies. This study also examines the interrelated issues of environmental legislation and coke plant closure and their impact on the spatial behavior of the coke industry. This study examined these relationships and changes in these relationships at the metropolitan areal scale from 1959 to 1989.;Based on the results of the primary empirical tests, the U.S. metallurgical coke industry consistently groups with such industries as the petroleum refining industry, the petrochemical industry, and the coal tar industry, in addition to the iron and steel industry at the urban and metropolitan areal scale. The empirical results also revealed that coke plants have greater rates of closures in air pollution control attainment areas than in air pollution control nonattainment areas. In addition, coke plant closures in general seem to be inversely related with respect to plant size and metropolitan areal size. Specifically, large coke plants will have a greater tendency to close in small metropolitan areas and small coke plants will have a greater tendency to close in large metropolitan areas.;Likewise, the empirical results show that although there have been a number of inter and intraregional shifts in coke plant locations over time, for the majority of the coke plants that presently exist, inertia has been a strong force in maintaining their present locations since the 1920s.;These empirical results were obtained through the use of correlation, regression, and contingency analyses. The primary data utilized were metropolitan area data from 1959 to 1989.