The EPA Facility Registry Service (FRS) is the massive digital archive of the United States Environmental Protection Agency. It lists over 800,000 sites: factories, landfills, treatment plants, and even research centers. Each site has a rich profile: address, activity type, hazardous substances handled, and especially the SIC code. The SIC, or Standard Industrial Classification, is a fourβdigit number that categorises a company's economic activity. Think of the SIC as a movie genre: it tells you whether the site is a steel mill, a power plant, or a recycling facility. Thanks to the SIC, analysts can quickly group sites by sector, compare emissions, and spot pollution hotspots. The FRS makes these data public, so researchers, journalists, and citizens can monitor the environmental impact of industrial activities. Moreover, regulators use the SIC to ensure companies comply with sectorβspecific rules. In practice, the SIC code is the key that unlocks a detailed map of the American industrial landscape, enabling more informed decisions for environmental protection.
The SIC was introduced in the United States in 1937 by the federal government to standardize the classification of economic activities. The code is structured in four levels: the first two digits indicate the main sector (e.g., 33 for machinery manufacturing), the third digit specifies the subcategory (e.g., 331 for metalβworking machinery), and the fourth digit identifies the most detailed class (e.g., 3312 for ironβworking machinery). The FRS uses the SIC to link facilities to specific environmental regulations, such as atmospheric pollutant emission limits or waste management requirements. Moreover, the SIC enables integration with other federal databases, like the Toxic Release Inventory (TRI) and the National Pollutant Discharge Elimination System (NPDES), facilitating crossβsector analyses. Researchers apply data mining and machineβlearning techniques to detect nonβcompliance patterns, leveraging SIC granularity to build predictive environmental risk models. This approach supports more targeted policies and timely interventions, reducing the environmental impact of industrial activities.
