Raceway/Tray Failures
Raceway/Tray Failures
Raceways (cable trays, conduits and other routes) do not fail on a single temperature the way components do. What matters is the cables they carry, so FRI3D works out a failure time for each cable and then decides whether the raceway has failed.
There are four ways a raceway or its cables can fail:
| How | Applies to | Failure time |
|---|---|---|
| Assigned to the fire source | The raceway and every cable it carries | Time zero |
| Heat Soak | Cables with no cable material, and the raceway itself | When the damage integral reaches 1 |
| THIEF | Cables with a cable material | When the cable's inside temperature reaches the THIEF failure temperature |
| FLASH-CAT ignition (CFAST) | Trays the FLASH-CAT fire reaches, and every cable they carry | When the flame front reaches the tray |
A raceway or cable that fails in more than one way keeps its earliest failure time, with the method that gave it.
Assigned to the fire source
A raceway assigned directly to a fire source fails at time zero, and so does every cable it carries, including cables that are also routed through other raceways. See Cables of Source-Assigned Raceways.
What the raceway is exposed to
Each raceway is a target in the simulation, measured on its face nearest the fire.
| Solver | Temperature | Heat flux |
|---|---|---|
| CFAST | Gas temperature around the target (TRGGAST) | Incident heat flux on the target (TRGFLXI) |
| FDS | Gas temperature off the face (<name>_TEMP), along its outward normal by the larger of 0.3 m and 1.5 times the grid cell size | Gauge heat flux at the centre of the first gas cell off the face, looking outward (<name>_HEATFLUX) |
Raceway failures use the gas temperature and the heat flux. The FDS wall temperature of the raceway is recorded but not used, because the cables, not the tray's surface, are what fail.
Heat Soak: the default
A cable left with no cable material ("Null") is evaluated with Heat Soak. Heat Soak accumulates damage over time from two NUREG/CR-6850 time-to-damage tables, one by exposure temperature and one by heat flux. At each moment it uses whichever of the two damages faster. The cable fails when the accumulated damage reaches 1.
| Cable type | Damage starts at |
|---|---|
| Thermoplastic | 205 °C or 1 kW/m² |
| Thermoset | 330 °C or 1 kW/m² |
The cable type for Heat Soak is the scenario's default cable type, thermoplastic unless changed. That is the conservative choice.
All heat-soak cables in a raceway share the raceway's exposure, so they fail at the same time.
THIEF: cables with a material
A cable given a cable material (dimensions and thermal data) is evaluated with THIEF. THIEF models the cable heating up and fails it when the temperature just inside the jacket reaches:
| Cable type | THIEF failure temperature |
|---|---|
| Thermoplastic | 200 °C |
| Thermoset | 400 °C |
- CFAST: FRI3D runs THIEF on the raceway's gas temperature history. A cable in a conduit that has a material is modelled inside the conduit's steel shell.
- FDS with FLASH-CAT: FDS runs THIEF itself, with a device at the cable, and FRI3D uses that result.
- FDS otherwise: FRI3D runs THIEF on the raceway's FDS gas temperature history, as for CFAST.
When the raceway fails
A raceway is marked failed when:
- all of its cables have failed, or
- it carries no cables and its own Heat Soak exposure fails it.
A raceway with some failed cables and some surviving ones is not failed, but its failed cables are. Each cable is judged on its own exposure and method, so a raceway that fails this way does not fail its cables. Only source assignment and FLASH-CAT ignition fail a raceway together with every cable it carries.
The raceway's failure time
The failure time recorded for the raceway itself is always its own Heat Soak time, and its failure method is always reported as Heat Soak, whichever method failed its cables. It is not taken from the cables' failure times. This matters when a raceway carries THIEF cables:
| Situation | Raceway's recorded failure time |
|---|---|
| Every cable is a Heat Soak cable | The Heat Soak time, the same as its cables |
| THIEF cables, and the raceway's Heat Soak also fails | The Heat Soak time, which can be earlier or later than the THIEF cables' times |
| THIEF cables all fail, but the raceway's Heat Soak never does | Time zero |
In the last case the raceway shows as failed from the start of the fire, even though its cables failed later. When you report a raceway that carries THIEF cables, use the failure times of its cables, which are worked out for each cable, rather than the raceway's own time.
Source assignment and FLASH-CAT ignition are the exceptions: when either gives an earlier time, the raceway is recorded at that time and with that method instead of its Heat Soak time.
FLASH-CAT: trays that catch fire
When FLASH-CAT is on in a CFAST scenario, the cable trays above the fire can ignite and the fire spreads along them. A tray that ignites fails at the moment the flames reach it, and so does every cable it carries. NUREG/CR-6850 uses the same thresholds for cable ignition and cable damage, so a burning cable has failed.
When a tray ignites
- The lowest tray over the fire ignites when the source fire's flames reach its underside. If they never do, it ignites when its exposure passes the ignition threshold: 205 °C or 6 kW/m² for thermoplastic cables, 330 °C or 11 kW/m² for thermoset.
- Each tray level above ignites 4, 3, 2, then 1 minute after the level below, as in FLASH-CAT.
- Along the trays, a length of cable centred over the fire ignites at once. From its ends the flame front moves along the trays at 3.2 m/h for thermoplastic cables (1.1 m/h for thermoset), in both directions.
- Connected trays. Trays that touch, or are linked, at the same level carry the fire on: the front passes from one tray into the next where they join, and continues along it. A tray ignites when the front reaches it by the shortest route along the trays.
So a tray directly over the fire fails within seconds of the flames reaching it, while a tray connected to it 2.5 m away fails only when the front gets there, about 40 minutes later for thermoplastic cables. In a short simulation, a connected tray may not ignite at all.
FLASH-CAT itself describes one stack of straight trays, with the fire spreading "until the end of the tray is reached" (NUREG/CR-7010). Carrying the fire on into connected trays is FRI3D's extension of the method.
The tray's probe can read cool while the tray is burning
CFAST does not model flames on the trays. It releases the trays' fire as one plume above the source and spreads its heat into the hot gas layer. A tray away from the source, or below the hot layer, can show a low probe temperature while FLASH-CAT has it burning. The ignition time is what fails the tray; the probe only shows the room around it. The probes do include the trays' fire, so the Heat Soak and THIEF results of every raceway already account for it.
FDS
In FDS the trays' fire is modelled on the trays themselves, and ignition is not recorded as a separate failure. Raceways and cables fail on their Heat Soak and THIEF results from FDS, which include the burning trays.
See FLASH-CAT for the heat release of the burning trays.
What's next?
- Component Failures: how equipment fails.
- Heat Soak and THIEF: the two cable failure methods in detail.