Carbon bed fires and the use of carbon canisters for air emissions control on fixed-roof tanks

Fixed-roof tanks are used extensively at manufacturing, waste management, and other facilities to store or process liquids containing volatile organic compounds. Federal and state air standards require the control of organic air emissions from many of these tanks. A common practice used for some fixed-roof tanks that are required to use controls is to vent the tank through an activated carbon canister. When organic vapors are adsorbed on activated carbon, heat is released. Under certain conditions, the temperature of the carbon bed can increase to a level at which the carbon or organic vapors spontaneously ignite, starting a fire in the carbon bed. Bed fires in carbon canisters are not uncommon and can present a significant safety hazard at facilities if proper safety measures are not implemented. This article discusses how carbon adsorber bed fires occur and presents general guidance on safety measures for carbon canisters installed on fixed-roof tanks to reduce the likelihood of a carbon bed fire and to minimize the impact in the event of a fire.     

Sensitivity Analysis of Dispersion Models for Point and Area Sources

The body of information presented in this paper is directed to air quality managers in industry and government contemplating modeling emissions from complex sources under the bubble concept.

Point and area source algorithms of PAL, RAM, and ISC-ST were analyzed to show the effect of various input assumptions on model output. Several important parameters were varied individually; receptor grid spacing, emission release height, area source size and source type. Each of these parameters was varied over a range of values while all other modeling parameters, both physical and meteorological, were held constant. The outputs of each model are plotted for easy comparison.

Results indicate that it would be inappropriate to make certain assumptions regarding source characteristics without knowing the behavior of each model. The graphs show how the model predictions can vary for different input parameters when applied to point and area sources. The paper presents general rules of thumb for evaluating model results for many applications such as the bubble concept, emissions banking, offsets, and new source reviews. The results serve as a guide in selecting and using models for both point and area sources.