Drexel Chemical Company - Executive Summary |
|
Drexel Chemical Company's (Drexel) accidental release prevention policy involves a unified approach that integrates technologies, procedures, and management practices. All applicable procedures of the U.S. Environmental Protection Agency (EPA) Prevention Program are adhered to. Drexel's evergency response policy involves the preparation of response plans which are tailored to each facility and to the emergency response services available in the community, and is in compliance with the EPA Emergency Response Program requirements. The Drexel chemical manufacturing facility located at the Tunica Industrial Park in Tunica, MS includes tank containers, production areas, process formulation areas, various safety equipment, and an instrument room which contains a flow recorder, various electrical parts, and additional safety equipment. The amount of Monomethylamine (MMA) handled is dependent on seasonal production demands, but at times exceeds the threshold quantity listed under EPA's Ris k Management Plan (RMP) requirements for a Program 3 facility. The facility is normally manned by 30 production, maintenance, and management personnel, according to seasonal production fluctuations. The offsite consequence analysis includes consideration of two MMA release scenarios, identified as "worst case release" and "alternative scenario". The first scenario is defined by EPA, which states that "the owner or operator shall assume that the... maximum quantity in the largest vessel... is released as a gas over 10 minutes," due to an unspecified failure. The alternative scenario is defined as "more likely to occur than the worst-case release scenario." Atmospheric dispersion modeling has to be performed to determine the distance traveled by the MMA released before its concentration decreases to the "toxic endpoint" selected by EPA of 3 ppm, which is the Emergency Response Planning Guideline (ERPG-3). This is defined by the American Industrial Hygiene Association (AIHA) as the "maximum airborne concentration below which it is believed that nearly all individuals could be exposed for up to one hour without experiencing or developing irreversible or other serious health effects or symptoms which could impair an individual's ability to take protective action." The residential distance has to be defined, "to estimate the population potentially affected." The worst-case scenario at the Tunica facility involves a failure of one twelve thousand gallon tank (greater than 10,000 lb. of MMA). The offsite consequence analysis for this scenario was performed for two sets of conditions. The first set followed conditions pre-defined by EPA, namely release of the entire amount as a gas in 10 minutes, use of the one-hour ERPG-3 as the toxic endpoint, and consideration of the population residing within a full circle with radius corresponding to the toxic endpoint distance. EPA set these conditions to facilitate the performance of the offsite consequence analysis; howeve r, the assumptions used may be unrealistic because: (1) Only a fraction of the compressed liquified MMA released to the atmosphere flashes as a vapor. (2) It is not appropriate to compare a 10 minute release to a one-hour average standard. The 3 ppm one-hour ERPG-3 value can be modified using available time of exposure/concentration relationships to match the 10 minute release time. (3) Only the population within an elliptical plume extending downwind of the release point is potentially affected. This plume area, or footprint, is approximately 6% of the area of the circle. EPA mandated meterological conditions, namely Stability F, wind speed of 1.5 m/sec, highest daily maximum temperature and average humidity were used for both sets. When atmospheric dispersion modeling for the worst-case scenario was performed using the EPA assumptions, a distance to the toxic endpoint of 0.2 miles and an estimate of residential population potentially affected (<100) was obtained. The alternat ive release scenario involves the rupture of the excess flow valve connected to the twelve thousand gallon tank, possibly due to an earthquake. The ammount of MMA released is 1,5000 lb., at an average rate over one hour (the duration of the release) of 100 lb/min. Toxic endpoint distanc to ERPG-3 level was obtained. the latter is defined by AIHA as "the maximum airborne concentration below which it is believed that nearly all individuals could be exposed for up to one hour without experiencing or developing life-threatening health effects." The typical meterological conditions used were Stability D, wind speed 3.0 m/s, average air temperature of 77 degrees. The estimated distance traveled to the toxic endpoints is 0.3 miles and the residential population afffected (<100) for the ERPG-3. Actuation of manual back-up valves is an active mitigation measure considered. An additional mitigation system installed is the dike containing the existing MMA storage tank and a supplied water s ource making available a water spray (fog) t reduce air borne release. This will reduce significantly the amount of MMA released in case of an excess flow valve rupture. The general Drexel accidental release prevention program is based on the following key elements: - High level of training of the process operators. - Preventive maintenance program. - Use of process and safety equipment. - Use of accurate and effective operating procedures, written with the participation of the operators. - Performance of a hazard review of equipment and procedures. - Implementation of a daily auditing and inspection program. Chemical specific prevention steps include availability of Personal Protective Equipment (PPE) worn by operators during connection/disconnection of the MMA supply, awareness of the hazardous and toxic properties of MMA, and presence of other safety equipment in case of accidental release. No accidental releases of MMA have occured at this facility. The facility has an emerge ncy response program, which has been coordinated (reviewed) by the City of Tunica Fire Department. This program includes an emergency response decision tree and a notification plan. Emergency response drills and drill evaluations are conducted; emergency operation and response procedures are also reviewed at that time. |