Wyman-Gordon Forgings, Incorporated - Executive Summary |
|
The Wyman-Gordon Forgings, Incorporated, Houston facility is a manufacturer of ferrous and nonferrous forgings. The facility serves such international markets as commercial and military aerospace and land based gas turbine and power generation. Processes conducted at the facility include chemical milling and etching. These processes utilize hydrofluoric acid (HF), nitric acid, and hydrochloric acid. However, only hydrofluoric acid, with a typical maximum storage quantity of 2,900 pounds, is maintained at a level above the threshold quantity and concentration, thus requiring the facility to submit a RMP. The facility will adhere to the general duty provisions for each chemical used in a process below the threshold quantity/concentration. The process of concern at the Houston facility is the transfer and storage of HF for use in the etching operations. HF is received at Wyman-Gordon's (W-G) Houston, Texas facility in 55-gallon drums. Drum containers of HF are stored as well as us ed in the acid etch building. W-G has an administrative procedure that requires that drum containers of HF be stored ONLY in the acid etch building. Upon delivery at the acid etch building, the drums are placed in a storage area of the building that is outfitted with a secondary containment system. W-G uses the 55-gallon drums of HF to provide make-up acid for acid etch process tanks. When acid is needed for a process tank, the acid is pumped directly from the drum in the storage area using a drum pump/transfer hose setup. The worst-case scenario involves the unmitigated release of the contents of one (1) 55-gallon drum of HF outdoors, such as might happen if a drum were to fall off a delivery or fork truck and rupture. This scenario assumes that the acid forms a pool with a depth of 1 centimeter and that the pool evaporates at 770F. RMP*Comp, Version 1.06 was used to perform the worst-case release offsite consequence analysis. The release rate was estimated using the procedu res in EPA's OCA Guidance for the evaporation of unmitigated releases of liquids. The estimated distance to the toxic endpoint, determined using the procedures in EPA's OCA Guidance, is 0.6 miles based on "rural" topography. The rural topography assumption is appropriate for this analysis since the facility buildings are not extremely tall and the area is surrounded by relatively flat terrain. Landview III was used to determine the number of persons living within the impact area based on census data. According to Landview III's census database, there are approximately 1100 people living within the worst-case impact radius of 0.6 miles. There are residences and major commercial, office or industrial areas within the impact radius. There are no other public or environmental receptors within this radius. The alternative case scenario assumes that a drum is pierced near the bottom with a fork from the fork truck indoors and that the entire contents (55 gallons) is released onto th e building floor, forming a 1-centimeter deep pool, and evaporates at 770F. The building air is exhausted to a scrubber system with a removal efficiency of 96%. However, the overhead doors of the building are normally left open, and since the scrubber system was designed to operate with the overhead doors closed, it was assumed that the efficiency of the scrubber system is reduced to 50%. Further assumptions are that the truck fork makes a 15 square inch puncture in the drum and that the puncture is 4 feet below the acid surface. RMP*Comp, Version 1.06 was used to perform the alternative case release scenario analysis using the calculations and assumptions contained in EPA's OCA Guidance. Rural topography was assumed in the analysis. The release rate was estimated by using the procedures in EPA's OCA Guidance for the evaporation of releases of liquids passively mitigated by containment within a building and actively mitigated by a scrubber system on the building exhaust: The es timated distance to the toxic endpoint is 0.1 miles. Landview III was used to determine the number of persons living within the impact area based on census data. According to Landview III's census database, there are approximately 30 people living within the alternative case impact radius of 0.1 miles. However, this estimate of population is based on the block group proration method, which is statistical in nature. In reality, a distance of 0.1 miles from the release point does not even reach to the Wyman-Gordon fence line, so the estimated population is zero and there are no public or environmental receptors within this radius. The facility has not had an accidental release of HF from this covered process in the five years prior to the submission of this Risk Management Plan (RMP). |