As new programs are initiated by industry and federal or state regulatory agencies to reduce releases of hazardous air pollutants and reactive volatile organic compounds into the atmosphere, there are increasing questions over the loss of organic compounds from fugitive sources, the best method to estimate these losses, and the factors that influence these losses. The presently established method is to use generalized emission factors for each specific piece of equipment comprising a process unit. The accuracy of the estimates obtained using the available emission factors depends upon two features. These are first, the precision and accuracy of the actual emission factors or correlations themselves, and second, the resemblance of the systems from which the correlations were developed to those for which they will be applied.
INTRODUCTIONSignificant fugitive emissions can occur due to design factors and inadequate monitoring or maintenance. To develop the existing industry-wide fugitive emission factors, the U.S. Environmental Protection Agency (EPA) sponsored an extensive sampling program and analyzed a large amount of data collected from various types of operating units. Based upon this analysis they determined that although much of the equipment in a process unit leaks to some degree, a small percentage with significant leaks contribute most of the equipment mass emissions from a typical unit. Following extensive public review and comment, the EPA evaluated the cost effectiveness of alternative control techniques for minimizing fugitive emissions, and defined a cost effective cutoff which would determine when a leak is of sufficient magnitude to warrant repair [ I ] . This cutoff, based upon a hydrocarbon screening measurement at the potential leak interface, was set at a concentration level of 10,000 ppm. In the absence of any data to the contrary, the present regulatory policy to apply these factors is to assume that leaking equipment of a specific type (i.e., gas valves, liquid valves, flanges, pumps, etc.) within a process unit tends to emit similar quantities of VOC as leaking equipment of the same type in another process unit [2]. Once a component is defined as leaking (i.e., above the 10,000 ppm cuton), it is assumed to have VOC emissions at a certain average mass rate. Similarly, if the equipment of a specific type is defined as 'nonleaking' (i.e., below the 10,000 ppm cutofn, it is assumed to have emissions at a certain lower average mass rate. The average leak rate for specific types of equipment in a process unit was determined using a composite of the leaking and non-leaking equipment [3]. These fugitive emission factors for leaking, non-leaking and average equipment were known as the SOCMI (Synthetic Organic Chemical Manufacturing Industry) Fugitive Emission Factors. Up to the present time, state and federal regulatory staff have been limited to using these factors to estimate fugitive emissions of organic compounds from typical process units in developing emission inventories or specific emission s...