API PUBL 337-1996

$39.00

Development of Emission Factors for Leaks in Refinery Components in Heavy Liquid Service

Published By	Publication Date	Number of Pages
API	1996	72

Guaranteed Safe Checkout

Category: API

If you have any questions, feel free to reach out to our online customer service team by clicking on the bottom right corner. We’re here to assist you 24/7.
Email:[email protected]

Description

I Estimating air pollutants from stationary sources is necessary for compiling emission inventories, determining emission fees, and meeting the conditions of various permits and compliances.Extensive field measurements have been made over the last four years to develop more accurate emission correlation and emission factor data for estimating emissions from leaking pipelinecomponents. For any petroleum industry facility employing leak detection and repair (LDAR) procedures, the component-leak emissions can be estimated using the recorded screening valuesexpressed in parts per million (ppmv) along with component-specific correlation equations approved by EPA and found on its Technology Transfer Network (TTN) bulletin board (EPATTN, 1995). For certain types of facilities where LDAR is not practiced, such as oil production fields and bulk terminals, EPA emission factors are readily available for use in estimatingemissions.

Pipeline components in heavy liquid (HL) service are generally not included in LDAR programs regardless of the facility. Since HL screening values are generally not available, it is necessary touse EPA’s emissions factors. The HL emission factors published in EPA’s 1993 Protocol for Equipment Leak Estimates (EPA, 1993) were believed to be high, based on available information.Therefore, API conducted a study to develop more accurate HL emission factors.

OBJECTIVE

The objective of the program was to develop the emission factors and to determine whether or not the type of distillate or residual hydrocarbon in the HL stream would influence the emission factors. Emission factor data were obtained and extensive statistical analysis of the screening data and related process parameters was performed and discussed.

APPROACH

Refineries in Southern California (SoCal) were solicited for available HL screening data. Four refineries responded, providing 2 1 1,000 discrete HL screening values. At those refineries,LDAR has been practiced for approximately ten years on components in gas (G) and light liquid (LL) service, but not generally for HL service. Nevertheless, because SoCal has tight emission controls, additional HL component screenings were conducted in an area without LDAR to determine whether or not the Soca1 HL screening values were representative. Two refineries in Washington State were chosen for the HL screenings. Washington is in attainment of the National Ambient Air Quality Standards and LDAR programs are not required except for new or modified facilities under the New Source Performance Standards (NSPS). More than 2,500 discrete values were recorded from which the average emission factors for each component category were computed. The Washington test used a sampling matrix which covered a representative range of middle distillate and residual process streams.

REPORT ORGANIZATION

Section 2 presents the SoCal data from four refineries showing leak rate distribution and average emission rates for various components. Section 3 presents the Washington State refinery data including the screening values, leak rate distribution, average emissions and the effects of stream composition and temperature on leak rate. Section 4 presents the conclusions. The statisticaltreatment of the SoCal, Washington and aggregated data is presented in Appendix A and the discrete screening and associated measurements, taken at the Washington refineries, are tabulated in Appendix B.

Additional information