Craig Station - Executive Summary |
|
INTRODUCTION Located near Craig, Colorado, Craig Station (the Station) is the largest coal-fired generating station in Colorado. Its net capacity is 1,264 megawatts. Coal is purchased from local coal mines and is transported to the site either by truck or train. The coal is burned to produce heat and generate steam. the steam is used to turn turbine-generators that produce electrical energy. The facility has three units which are conventional coal-fired electric generating units. Units 1 and 2 are each capable of generating 428 megawatts, net output. Unit 3 is capable of generating 408 megawatts, net output At the Craig Station, coal is crushed into powder and blown into massive boilers, where it is burned to heat water and produce steam. The forced steam spins the turbine blades in much the same way a breeze turns a windmill. The turbine rotates a shaft that drives the generator. The generator's magnet turns inside a stationary magnet wrapped with copper wire and, as the magnet rotates, electric current is produced on the wire. Craig Station has three different generators: Unit 1, Unit 2, and Unit 3. Water is used to condense steam exhaust from the turbines. The water used for cooling is cooled in a tower by forcing ambient air through falling water droplets. Water used in the cooling towers at the Craig Station must first be disinfected with chlorine to control the growth of algae, slime, and other organic life. The use of chlorine as a biocide increases the efficiency and lifetime of the condensers and the cooling towers. Two chlorination systems are in place at the Craig Station: the Unit 1 & 2 Chlorination System and the Unit 3 Chlorination System. THE UNIT 1 & 2 AND UNIT 3 CHLORINATION SYSTEMS The Unit 1 & 2 System provides gaseous chlorine under vacuum directly from regulators mounted on each of six one-ton containers. Three of these containers have been designated for Unit 1, and the other three for Unit 2; however, a chlorine feed cross tie is included so that each header can feed either one or both rate controllers. The Unit 3 System is similar in design to the Unit 1 & 2 System. Both systems use vacuum regulators to draw gaseous chlorine from one of six one-ton chlorine containers. Because the Unit 3 System supplies chlorine to only the Unit 3 process water, it does not include the cross tie pipe section and block valve which is located in the Chlorine Control Room of the Unit 1 & 2 System. Additionally, the Unit 3 System requires only one rate controller, and container change-outs are not as frequent as for the Unit 1 & 2 System. The materials of construction, principles of operation, and maintenance requirements are generally the same for both systems. Because the maximum possible inventories of each of these processes (24,000 lbs) exceed both the EPA RMP threshold quantity (2,500 lbs) and the OSHA Process Safety Management (PSM) threshold quantity (1,500 lbs), they are subject to the Program 3 requirements of the RMP regulations. Program 3 requires an offsite consequence analysis of the worst-case and alternative release scenarios, a Prevention Program (also required by PSM), and an Emergency Response Program. WORST-CASE RELEASE SCENARIO The worst-case scenario is based on the complete failure of the largest vessel of a chlorination system in a 10 minute period (a single ton container). The analysis requires the calculation of the area surrounding the release point that is subjected to chlorine concentrations greater than or equal to the toxic endpoint. The toxic endpoint (EP) has been set at the ERPG-2 (Emergency Response Planning Guideline) concentration, which is defined by the American Industrial Hygiene Association. This EP is equal to the maximum airborne concentration below which it is believed that nearly all individuals could be exposed for up to 1 hour without experiencing or developing irreversible or other serious health effects or symptoms that could impair an individuals ability to take protective action. For chlorine, the EP is 3 parts per million (0.0087 mg/L). Both of the chlorination systems manifold six containers together; however only one container per process is valved in at all times. Therefore, the worst-case scenario for each process involves the entire contents of a one-ton container being released over a 10 minute period. Because the chlorine containers at the facility are located in an enclosed building in direct contact with the outside air, the RMP Offsite Consequence Analysis Guidance suggests using 55% of the release rate calculated for the same scenario occurring outside the building. EPA's RMP*Comp program defines the distance to the toxic endpoint for this scenario as 2.2 miles.(It should be noted that the two systems are approximately 1000 feet apart and are located in separate buildings. The consequences of a release from either process are identical). Some residences are located in the radius, with an average population of 32 people ( 1990 Census data--Block Group Proration Method). Two recreational areas are located within the worst-case impact radius: the motocross area and the Yampa River Boat Input, which is also a State Park. One additional non-residential public receptor is the Moffat County Landfill. No other recreational areas or other non-residential public receptors (e.g., hospitals, schools, etc.) are within these impact areas. Although the worst-case consequence analysis is required by the RMP rule, a worst-case occurrence is a highly unlikely event. Catastrophic failure of a chlorine ton container is extremely remote because: the chlorine process areas are monitored with a detection system that alerts plant personnel at the main part of the facility (the chlorination systems are located in separate buildings away from the main plant); and the external surface of the each chlorine container is visually inspected regularly before hook-up and during operation for any departures from normal. Containers t hat do not meet the specifications listed in the Cylinder and Ton Container Procedure for Chlorine Packaging (Chlorine Institute, Pamphlet 17, June; 1994) are returned to the supplier. Additionally, safe work practices and Standard Operating Procedures are employed during the change-out of chlorine containers. ALTERNATIVE-RELEASE SCENARIO The alternative-release scenario is considered more likely to occur than the worst-case scenario. A credible alternative-release scenario for chlorine was chosen after a review of previous hazard assessments and/or Process Hazard Analyses, facility or industry incident reports, maintenance work orders, incident recall, technical judgment and the Compliance Guidance and Model Risk Management Program for Water Treatment Plants scenarios (American Water Works Association). The scenarios discussed in the Compliance Guidance and Model Risk Management Program for Water Treatment Plants were considered the most appropriate resource for use in defin ing the alternative release scenario. Each scenario was evaluated and ranked based on severity and likelihood of occurrence. The release scenario chosen for analysis is the following: tubing failure, bad connection, or container valve failure resulting in the release of gas through the 5/16-inch-diameter valve body opening (CLA-3). It is assumed that the release continues for 60 minutes before it is remediated. The total amount of chlorine released is specified as 317 pounds (average rate = 317pounds/60 minutes = 5.3 lbs/min). This scenario was modeled as a release from a horizontal cylindrical tank with the release occurring through a short pipe or valve in the top of the tank, and the chlorine is assumed to escape only as a gas. Additional assumptions include rural surroundings, Atmospheric Stability Class D, wind speed of 6.7 miles per hour, relative humidity of 50%, and a temperature of 77 0F. Entering the release rate specified by this guidance into the RMP*Comp program and enteri ng the EPA-approved mitigation factor for the scenario occurring indoors (so that only 55% of the specified release rate is used in the calculation), the distance to the EP was calculated as 0.6 miles. The 1990 Census data (Block Group Proration Method) estimates that 2 people reside within the alternative-case impact area for chlorine releases. LandView III calculates populations within defined impact areas using two methods: the Block Group Uniform Population Density Method and the Block Group Centroid Method. The first method tallies data for each Census Block that has any portion within the defined radius of impact. It then prorates the results based on the ratio of the portion of the Census Block that is actually inside the radius of impact. This method distributes the people in the Census Block evenly across the land area and counts only those assigned within the radius of impact. This method is useful when only small portions of the Census Block are located within the radius of impact. The second method, the Group Centroid Method, finds all Census Blocks that have their centers inside the defined radius of impact. The populations of these blocks are added together, regardless of whether the entire block is inside the radius of impact or not. This method is most useful with impact radii greater than 1 mile in densely settled areas. Because the alternative-case impact area for a chlorine release at the Craig Station is relatively small, neither of these population estimation methods provide accurate results. The Group Centroid Method fails to provide any data because there are no centroids located within the radius of impact. The Block Group Uniform Population Density Method estimates two people within the worst case radius. However, surveys of the area surrounding the facility have shown that no residences are located within a 0.6 mile radius of the chlorination systems. The method is more useful for larger radii (e.g., those used for the worst-case rel ease scenarios). Based on drive around surveys, it is appropriate to assign zero population within the chlorine alternative-case impact radius. There no recreational, or institutional areas or environmental receptors located within this radius the alternative-release scenario. Although an alternative-release consequence analysis is required by the RMP rule, a failure and release that is large enough to reach the EP beyond the fence line is not likely because (1) the facility has a preventive maintenance program in place to maintain the ongoing integrity of the system, (2) Standard Operating Procedures and training programs are developed, implemented and maintained to ensure that the operations and maintenance personnel will avoid activities which may cause releases, and can quickly respond to any releases that do occur, and (3) preventive maintenance schedules are developed to ensure that process equipment operates optimally and safely. FIVE-YEAR ACCIDENT HISTORY Craig Station has not had an accident involving the chlorination systems during the past five years which resulted in any deaths, injuries, or responses or restoration activities for an exposure of an environmental receptor. Craig Station has never had an off-site release of chlorine. PREVENTION PROGRAM Tri-State has developed a combined Process Safety Management and Risk Management Program for the chlorination systems at the Craig Station. Because many requirements of the RMP and PSM rules are similar, an integrated approach is useful to ensure compliance with both rules, to eliminate redundancy, and to simplify documentation. The Prevention Program includes: a Management System, an Employee Participation Plan, Process Safety Information and Process Hazard Analysis requirements, Standardized Operating Procedures, a Training Program, a Contractor Safety Program, a Pre-Startup Safety Review requirement, a Mechanical Integrity Program, a Hot Work Permit System, a Management of Change Procedure, an I ncident Investigation Procedure, and an Audit Program. EMERGENCY PLANNING, PREVENTION AND RESPONSE PROGRAM Both the EPA RMP and OSHA PSM rules rely mainly on previous requirements for emergency action plans and training (29 CFR 1910.138 [a] and 1910.120[a], [p], and [q]). The Craig Station has consolidated these requirements into the Integrated Contingency Plan (ICP). In the same way that the PSM and RMP rules are efficiently satisfied by one program, the ICP will allow Craig Station to effectively manage its Emergency Planning, Prevention, and Response efforts while eliminating redundancy and excessive documentation. The ICP covers all regulatory requirements for emergency response training, evacuation procedures, procedures for notifying the public, government agencies, and local emergency response agencies of a release, procedures for the maintenance of emergency response equipment, emergency management and organization structures, and guides for conducting drills and table-top planning exercises. Craig Station is committed to ensuring the safety of its workers, the surrounding community, and the environment. The Station maintains its own Hazardous Materials Emergency Response (Hazmat Team), and several employees at the facility are also members of the local Fire Department and are trained as Emergency Medical Technicians. Additionally, the Craig Rural Fire Department conducts training drills on the Station property. Tri-State is also a member of the Local Emergency Planning Commission (LEPC). To learn more about Tri-State Generation & Transmission Association, Inc., Craig Station, and other Tri-State facilities, please visit the web site at http://www/tristategt.org. |