(65b) Fire Hazards at Power Generation Facilities

BACKGROUND Power generation plants present several challenging hazards. Some are well understood (e.g., rotating equipment) while others remain poorly defined or understood (e.g., lubricating oil systems, control oil systems, and hydraulic oil systems). In an effort to better define the hazards created by pumped oil systems for our clients and to help us develop better protection options, FM Global initiated a research project to investigate oil fire hazards expected at a power generation plant without an operating floor. FM Global loss statistics have shown the majority of fire loses within power generation facilities are due to mechanical failures and human error. These failures have resulted in numerous oil fires including spray fires, three-dimensional fires, and pool fires, and have involved lubricating oil, control oil and seal oil systems. A fire generated by flowing oil, either as a spray or three-dimensional spill, presents a very challenging situation. Water-based automatic sprinkler systems have limited effectiveness against these types of fires. The only way to extinguish spray or three-dimensional spill fires is to stop the oil flow; however, suddenly shutting down any of the oil systems on a turbine can result in severe damage to the equipment. Turbine operators generally are resistant to shutting down oil systems under any circumstances and turbine coast down times can be as high as 45 minutes. Therefore, a multi-level solution is required. Fire protection measures must be taken to minimize the impact of the fire to the facility while the oil is flowing, and operators must look for opportunities to shut down oil systems or at least reduce pressures and flow rates. OBJECTIVES The objectives of the project were to 1) provide detailed visual documentation of the potential fire scenarios associated with turbine halls within power generation facilities without an operating floor and 2) examine new and existing protection schemes to mitigate losses due to fires. A final project was aimed at defining a functional spray shield that could be used to mitigate spray fires at flanges. To meet these objectives a large-scale mock-up of a turbine unit was constructed under an 18.3 m (60 ft) high ceiling within the large burn laboratory at the FM Global Research Campus. The mock-up consisted of a lube oil tank, tank containment area, and the high- and intermediate-pressure sections of a turbine, located on a pedestal. A total of 22 large-scale tests were conducted. The test study consisted of 11 pool fires, 9 spray fires, and 2 three-dimensional spill fires. Mineral oil with a closed cup flash point of 140.6°C (285°F) was used for all of the tests. The most recent work looked at developing testing criteria for spray shields. The goal of a spray shield is to eliminate the momentum from a high velocity liquid release. This would not stop the leak but simply change its characteristics from a strong directional flame to a weaker 3-D spill fire. Shields need to be evaluated on their effectiveness in meeting this goal as well as their performance under severe fire exposures. Many commonly used spray shields would not meet both of these criteria. CONCLUSIONS This test series provided a clear confirmation of the severity of oil spray and three-dimensional spill fires and the challenge of providing effective fixed protection for the tested scenarios. In addition, further data on extinguishment of pool fires using water-based automatic sprinkler protection and local protection schemes for spray fires were developed. It was also shown that a spray shield can be improperly designed. Without testing, a spray shield may not remove the momentum of a spray leak; instead they may redirect the spray.