Setting a Permit Limit through Process and Emission Data Analysis – a Retrospective Federal Case

摘要

Firing an indurating furnace with natural gas alone to make fluxed pellets results in highest emission of nitrogen oxides (NOx) in taconite pellet production. Co-firing with an alternative fuel (biomass, distillate fuel oil, or coal) or making standard (“acid”) pellets reduces NOx emission. We are interested in setting a long-term NOx emission limit for a retrospective PSD permit. The NOx permit limit must be set so that takes into accounts a wide variety of fuels and operating conditions as they evolved over the past 10 years (1994 – 2003). The process and emission data over the 10-year period were fitted with a multiple linear regression model. With a set of input data derived from the process data (average values of rolling average data) in the most recent three years, the regression model generated rolling average values (3-month, 6-month, or 12-month averaging time) of predicted NOx emission that were used in setting the long-term permit limit. A simplistic check was performed to validate the predicted NOx emission values. The selected long-term NOx permit limit added enthusiasm to all parties in moving the permitting process forward, amid the struggle with Class I area impacts and other aspects of the permit.