There are two fire protection issues for Hall B; protection of the experiment hall and support facilities, and protection of the experimental apparatus. Of these, the experiment hall and support facilities present no fire hazards which cannot be addressed by compliance with the applicable codes and standards. The experimental apparatus, however, presents a fire protection problem which is simply not addressed by conventional codes and standards. For example, use of riser-rated cables for vertical cable ru ns is not possible because of the technical requirements for high-frequency transmission in the cables. A second example is the requirement for having an absolute minimum of material between the target and the detectors; this is crucial to their operation , and this consideration has dominated the design of the spectrometer as a whole. This precludes, however, the use of sprinklers or fire barriers to protect the detectors, which introduces a problem in that at least one detector system (the time-of-flight counters) consists of a large mass of fuel which is in no sense self-extinguishing. The problem is compounded by several other factors, such as the large inventory of cable plastic which is distributed throughout the hall, the sensitivity to smoke and wa ter damage of many of the detector and electronic components, and the one-of-a-kind nature of all of the detector systems. The impact of a major fire on the experimental facility would be large, both in terms of the dollar cost of replacing components and in terms of the delay to the experimental program.
Broadly speaking, there are two distinct types of fire scenarios for Hall B. The first is a smoldering cable fire occurring in one of the electronics areas; this is the most likely type of fire for Hall B. In this scenario the fire spreads relatively slow ly because the cables are fire-rated, and the cable runs are horizontal ( See Also: Fermilab Horizontal Cable Tray Fire Test Results ). The main damage in this case is due to production of smoke and to the extinguishing agent used to put out the fire. This type of fire is the most probable because it involves fuels in quantity which are constantly in the presence of credible ignition s ources. The harmful effects of the smoke damage should not be underestimated; for electronic circuitry the damage is primarily due to the corrosive nature of the combustion products of burning PVC. A smoldering cable fire could also occur in one of the ca ble storage rooms, but this is much less likely since there are no credible ignition sources during remote operations.
The second type of fire scenario is the `runaway' fire which spreads rapidly and which is limited in growth only by the available fuel supply. In Hall B, this scenario is less likely since the vulnerability to it is for those fuels which essentially do no t have a credible ignition source during remote operation. This includes the vertical cable runs on the forward carriage, and the time-of-flight counters. For the vertical cable runs, the damage from this type of fire could include a very large amount of smoke generation and extensive damage from an extinguishing agent, in addition to radiant and convective heat damage of detectors and electronics nearby. For the time-of-flight counters, this type of fire would produce extensive smoke damage, damage from extinguishing agents, direct loss of detector components, and heat damage to other detector and magnet components. Compounding these problems is the lack of access to these detectors in the installed position, where a straight line-of-sight would not be a vailable to allow spraying extinguishing agent directly on the fire.
In addition to these two basic fire scenarios, there are a number of other scenarios as detailed in See Also: Specific Fire Scenarios - Hazard Analysis and Risk Estimates .
The basic fire protection approach for Hall B includes the following elements:
Since conventional fire protection methodologies (sprinklers, multi-hour fire barriers, riser-rated cables, etc.) cannot be applied in several locations of the apparatus, a much higher emphasis has been placed on three elements of this approach: the redun dant, position-sensitive early warning systems, the centralized notification system, and the rapid response capability. To a large extent this extra emphasis compensates for the lack of more conventional protection components. It is also appropriate to co ncentrate on very early-stage detection because of the high cost of any programmatic delays in the experimental operation of the detector. If the early warning systems can be made sufficiently sensitive and the response to them is sufficiently prompt, any fire can in principle be extinguished before it causes significant damage.
In normal data taking mode, no one is in the hall. During controlled accesses, no more than ten people can be in the hall at any time (there are ten access keys); typically only two to four people are in the hall during these times. While there is a large variation, for the purposes of estimating occupancy, about one hour out of 24 can be considered to be devoted to controlled access, by an average of three people.
During maintenance periods the hall occupancy varies greatly. During working hours, an average number may be estimated to be ten. After working hours, the number may be estimated to be one person. At peak times, there have been as many as 25 workers in th e hall during working hours, for a few days. During public tours, there are typically 10-15 people in addition to workers, and there can be more.
In summary, taking into account the fraction of time the hall is in each operating mode (see See Also: Modes of Operations/Description of Facility Use ), the average occupancy is estimated in See Also: Occupancy estimates for Hall B. .
There are three exits to Hall B. The personnel access labyrinth door is at the floor level of the hall on the east side of the building; the truck access tunnel is at the floor level on the southwest side of the building. The third exit is through the ups tream beam penetration into the hall, and this is accessible from Level 1 of the space frame.
There are multiple exits from each Level of the platforms. In most cases there is a `normal' exit and an `emergency' exit, such as a vertical ladder. The exits are listed in See Also: Exits from each elevated platform level in Hall B. .
The maximum travel distance to get to a hall exit is approximately 160 feet. This distance assumes a worker is on the south end of Level 3 of the forward carriage and exits via either the truck ramp tunnel or the personnel access door. This number compare s favorably with, for example, the NFPA 101 requirements for general industry (maximum of 200 feet, unsprinklered, or 250 feet, sprinklered) or for special purpose industry (300 and 400 feet, respectively) 6 . Since, however, the truck access tunnel does not qualify as a fully protected exit, this number effectively depends on which exit is chosen. See the comments on smoke escape hoods and protection of means of egress in the following sections.
Of the three exits to the hall, two are to areas which are isolated by commercial smoke and fire barriers; the third, the truck ramp, is somewhat isolated if the rollup door is closed, and not isolated if it is open. Even if closed, the rollup door is not a rated fire barrier.
None of the exits from the platform Levels are protected from fire or smoke by a rated fire barrier. In case of a fire incident, there is effectively some smoke and convective heat protection in the early stages of the fire because the smoke is free to ri se to the dome (as demonstrated by one historical small fire). The large volume of the hall provides a buffer not typical of the average apartment building or office space. If the exiting occupant is close to the point of the fire, of course, this effect does not offer any protection.
Emergency lighting has been installed on all platforms. These were tested by Jefferson Lab EH&S staff after installation, and are maintained by plant services on request.
Because there are significant smoke and fire hazards in Hall B, and to partially compensate for the long maximum distance of travel and the unprotected truck access ramp, smoke escape masks have been installed in Hall B. These devices provide a minimum of 15 minutes of breathable air, with a much longer maximum time depending on severity of smoke conditions and degree of exertion of the user. They also provide protection to the head and neck of flash temperatures of up to 1200 degrees F for a duration of a few seconds. The particular brand of smoke mask used satisfies the European standard EN403 for `escape hoods'. (At the time of purchase there did not yet exist a North American standard for smoke masks.)
The smoke masks have been mounted at the exits to each platform, and in addition they have been mounted at locations where egress is particularly constrained. To use them the wearer opens the box or pouch, removes the plugs from the filter (the plugs are attached to the box), pulls the mask over the head and pulls a strap to tighten the mask against the head. Glasses do not have to be removed; long hair should be pulled into the mask. Although the operation is simple, proper training in the use of these d evices is needed.