This is an old revision of the document!
The hazards from SF6 are located in 2 locations:
The accelerator tank is filled with SF6 at 480 to 650 kPa (70 to 90 PSI). There is a blower, some filters, some dryers, and some piping connected to the tank to keep the SF6 circulated, clean, and dry. It is possible for any of this piping or the tank itself to leak SF6. Since SF6 is significantly heavier than air, it will flow to the lowest point, and gather there, until mixing with the air and dissipating. The pit under the accelerator is the lowest point in the lab. The O2 monitor can be seen from the top of the stairs leading into the pit. A plot of the oxygen readings from the pit can be found at https://inpp.ohio.edu/~oual/tandem/oual_Pit_O2.jpg
The storage tank and compressor for the SF6 are located in the loading dock. The amount of SF6 stored in the tank is generally much less than what is stored in the accelerator tank. When the tank is open to atmosphere, all of the SF6 is in the storage tank on the dock. The O2 monitor is located approximately 1.5m (5 ft) to the right of the door leading from the lab into the dock. A plot of the oxygen readings from the dock can be found at https://inpp.ohio.edu/~oual/tandem/oual_Dock_O2.jpg
The lab/office space in room 115 is where dewars of Liquid Nitrogen (LN2) are stored. Although air is composed of 78% nitrogen, LN2 is roughly 650 times more dense than air. A spill of LN2 can easily displace enough oxygen from a room to create an asphyxiation hazard. The O2 monitor in this room is located on the wall to the right of the door from the lab hallway, approximately 1m (3 ft) from the door.
Below this point is the nitty-gritty details on the oxygen monitors, wiring, manuals, etc. Operators shouldn't need to refer to this information, but it is made available for accelerator staff.
Oxygen deficiency is monitored in the lab using 2 AMI Model 221R Oxygen Deficiency Monitors. These monitors were installed in June 2021 by Don Carter and Greg Leblanc.
Normal oxygen content is 20.9%. The current oxygen reading will be be displayed on the red LED numbers on the monitor. The monitor should also show a green power LED under normal conditions.
If low oxygen is detected, an alarm will sound, and the area should be evacuated, and accelerator staff contacted.
Correspondence with AMI can be found at: ami_sensor_replacement.pdf
The monitors are equipped with remote probes, to allow placement closer to the floor, since the risk is primarily from heavier than air gasses. The Ziurconium Oxide sensors used in these devices are expected to have a 10 year service life. The sensor is located in the remote probe, and could theoretically be sent back for repair separately from the head unit. AMI recommends returning both the head unit and sensor together. : We should find out how to tell when the sensors might be failing.
The oxygen monitors come equipped with 3 modes of external communication: RS-485 (single-point), 4-20 mA analog output, and 2 relay outputs.
Communication with the RS-485 port on these devices is accomplished by using an FTDI USB-to-RS-485 adapter. This adapter only allows communication with a single device. Control of the devices is accomplished using the AMI command center software, running on Windows. This software is currently installed on a computer in the control room, but will eventually be migrated to the VESDA monitoring computer. Currently a single twisted pair line is connected from the control room to the unit in the vault, and then daisy-chained to the unit on the dock. However, since the software is only capable of communicating with a single unit, the connector must be unplugged from one unit or the other in order to successfully communicate.
These devices utilize a standard 4-20 mA analog signal for integration with control systems. A 510Ω 1% metal film resistor is installed in across the terminals, converting this to a 2-10V signal. 2 twisted-pairs are used from the control room to the pit, and then an one twisted-pair is used out to the dock.
These units are equipped with 2 alarm output relays, each with NO and NC contacts. Currently these outputs are not connected anywhere. We may eventually want to connect them to a “remote” set of alarms located in the control room.
Data logging for these devices is accomplished by landing the 2-10V signal wires into the HP/Agilent/Keysight 34980A unit located in the control room. : I would list the input numbers here, but I can't remember what they are (hopefully Don has some notes).