The Atlanta Coca-Cola Bottling Company - Executive Summary
Facility Overview |
The Atlanta Coca-Cola Bottling Company operates a soft-drink bottling facility in College Park, Georgia. The facility uses anhydrous ammonia in support of daily operations. Anhydrous ammonia is used for refrigeration and is present in the compressor, evaporator, and pressurized receiver(s) that comprise a closed-loop refrigeration system.
An evaluation of facility chemical use and inventory indicated that RMProgram applicability was limited to propane (flammable substance) and ammonia (toxic substance). Ammonia is used in the refrigeration system at College Park. The ammonia travels from the discharge of the compressors through the evaporative condensers to the high-pressure receiver. The refrigeration system is a closed-loop system. Approximately 2,000 lbs of ammonia are added to the system annually. Ammonia is received via tank trucks. Transfer hoses and associated equipment is connected and the unloading process is initiated. Ammonia
is transferred from the tank trucks to a high-pressure receiver tank located outside of the building. The total refrigeration system capacity is 12,150 lbs of ammonia, and the actual operating charge is 10,520 lbs. This exceeds the RMProgram threshold for ammonia of 10,000 lbs.
As part of the RMProgram, a Hazard Assessment of potential accidental release scenarios for College Park was conducted. The goal of the Hazard Assessment was to complete an OCA which includes assessment of the WCS as defined by the established parameters in the RMProgram regulation followed by the consideration of a more likely alternative release scenario (ARS). For toxic substances, the Hazard Assessment requires that the WCS and ARS consider the effects of a release down to a toxic endpoint. Toxic endpoints have been established in the RMProgram regulation for each listed substance. The toxic endpoint for ammonia is 0.14 milligrams per liter (mg/l). Although, EPA has developed a RMPCom
pTM program to simplify the Hazard Assessment, College Park utilized the ALOHA program developed by EPA, National Oceanic and Atmospheric Administration (NOAA), and National Safety Council to model the off-site impact for the worst-case and alternative release scenarios. The ALOHA program was chosen because it can simulate a dense gas, can simulate two-phase flow, and can simulate several release scenarios.
Worst Case Scenario (WCS)
Based on the RMProgram regulation, the input required for the worst-case scenario is the worst-case release amount, release time, specified meteorological parameters, and the toxic endpoint. For the ammonia system, the worst-case release amount is the quantity of the single largest vessel (10,520 lbs). Because ammonia is a gas at ambient temperature, the worst-case release amount (10,520 lbs) is completely released in a 10 minute period as specified in the RMProgram regulation for toxic gases. The release of 10,520 lbs results in a release rate of 1,05
2 lbs/min. As required by the RMProgram regulation, the meteorological conditions for a WCS are a wind speed of 3.4 miles per hour (mph) and F stability class.
For this extremely conservative WCS, the distance to the toxic endpoint for the WCS is 1.8 miles that encompasses an affected area of 10.2 square miles. In order to determine the impacted population from the worst-case scenario, the Landview III program was used. Landview III is an information system that contains EPA regulated sites and demographic and economic data from the Bureau of Census presented in a geographic context. It is estimated that a population of 22,800 is located within the affected area. Public receptors within the impact area include residential areas, public recreation areas, schools, commercial/industrial areas, and Hartsfield International Airport. There are no environmental receptors affected.
Alternate Release Scenario (ARS)
As determined by the process hazard analysis (PHA) under the Accident Pr
evention Program, the ARS for the ammonia process is an outdoor release of ammonia liquified under pressure from a pump seal leak or gasket rupture. It was assumed that the release was mitigated within ten (10) minutes of the initiating event. No credit was taken for passive mitigation, since the release would not be in a contained area.
The liquid release is 1,340 lbs, with a release duration of ten (10) minutes, resulting in an release rate of 134 lbs/min. The ARS meteorological conditions are a wind speed of 7.9 knots and "C" stability class. In order to determine the typical wind speed and temperature for the ARS, meteorological data for Station 13874 (Atlanta) for 1992 and 1993 were obtained from the National Oceanic and Atmospheric Administration in Asheville, North Carolina. The modal or prevalent temperature (62 oF) and wind speed (7.9 knots) were used as the typical weather conditions for the ARS.
For the more realistic ARS, the distance to the toxic endpoint for the ARS
is 0.13 miles which encompasses an affected area of 0.05 square miles. In order to determine the impacted population from the ARS, the Landview III program was used. A population of 150 would be affected by the release. Public receptors included residential and commercial/industrial areas. There are no environmental receptors affected.
Accidental Release History
A thorough review of the College Park facility records in conjunction with discussions with environmental personnel revealed that there have been no known ammonia or propane releases in the past five years from College Park.
Accident Prevention Program
The facility complies with the Accident Prevention Program component of the RMProgram. The facility has compiled and maintains safety-related information for the regulated substances and processes, has completed a PHA for the applicable substances and processes, has developed operating procedures, provides employee training on the operating procedures, has developed ma
intenance procedures for the process equipment, has established routine compliance audits for the facility, and has developed an incident investigation protocol. All information required by the Accident Prevention Program is maintained continuously and is immediately updated if a major change occurs.
The required safety information for College Park has been compiled including a material safety data sheets (MSDS) for ammonia; and a summary table which documents the maximum inventory of ammonia storage at College Park, safe temperatures and pressures for ammonia use and storage, process equipment specifications, and process design codes.
A PHA has been performed for the ammonia process at the College Park facility as part of their PSM program. The PHA identifies the hazards associated with the process and examines potential accidental release scenarios. As a result of the review adequate safeguards to prevent releases must be developed and a relea
se detection system must be implemented. The PHA was performed using the What If? PHA method. The What If? experience-based technique involves the use of interrogating questions about equipment, operator, instrumentation, utility, or procedural failures in a structured form. The results of the hazard review were used to define the ARS for the Hazard Assessment (Attachment 4).
Accident Prevention Program
As required by the RMProgram regulation, the facility has developed and implemented accident prevention procedures including operating and maintenance, training, management of change, pre-startup review, employee participation plan, hot work permit, compliance audits, incident investigation and contractor safety management for the ammonia process.
Emergency response actions are coordinated with the local fire department. An emergency response plan has been prepared for the College Park facility which outlines emergency response actions to releases of ammonia,
internal and external notification/communication, emergency health care, and training.