# Style Guide for Word Users for the NIST Special Publication Format

Document Sample

```					                                             Chapter 1
FIRE MODELING

1.1         OVERVIEW
This chapter describes the numerical simulations of the fires within WTC 7. These calculations are very
similar to those conducted for WTC 1 and 2 (NIST NCSTAR 1-5F), and therefore only a brief description
of the major assumptions is included.
Fire simulations of WTC 7 were conducted for Floors 5, 7, 8 and 12. Except for 5, these floors were
chosen because a reasonable timeline could be reconstructed from the photographic evidence and
eyewitness testimony. However, there are far fewer photographs of WTC 7 than the towers; thus, it was
not possible to do as detailed an assessment of the fire. The observed fire patterns on Floors 11 and 13
were similar to those of Floor 12; thus, the simulation of Floor 12 is taken as representative of all three
floors, with appropriate offsets in time.

1.2         MODEL INPUT

1.2.1       Numerical Grid
Unlike WTC 1 and 2, the fire simulations were conducted for each floor individually, with no connections
floor to floor. Figure 1-1 displays the model of the 8th floor, showing the partitions and workstations. The
numerical grid for each floor of WTC 7 was the same as for WTC 1 and 2. Grid cells of 0.5 m by 0.5 m
by 0.4 m (vertical) were used, with a slab to slab height of 3.6 m. Although this grid is relatively coarse
for a computational fluid dynamics (CFD) calculation, it was determined from validation experiments that
it could adequately capture the peak temperatures, heat release rate, and burning duration of fully-
engulfing workstation fires. The coarse grid would not be appropriate to describe the very early stages of
the fires on each floor, nor would it help determine how the fires started or how the fires might have
spread floor to floor. Much of this is unknown because it occurred in the damaged and heavily smoke-
shrouded southern portion of the building. For this reason, the simulated fires were specified as having
originated near the southern face of the building with a size of about 2 MW. This is roughly the
equivalent of a single office workstation burning. As the fires grew and spread to other workstations, the
coarse grid did reproduce qualitatively the observed fire activity.
Appendix B                                                  Confidential and Predecisional Document, 1/11/2009

Figure 1-1. Model of the 8th Floor, WTC 7.

1.2.2        Floor Layouts and Combustible Load
Once the spatial resolution of the numerical grid was decided upon, architectural drawings were needed at
a level of detail consistent with the underlying numerical grid.
The furnishings in most of the open areas of WTC 7 were assumed to have consisted of office
workstations similar to those in WTC 1 and 2. Figure 1-2 shows a typical office environment in the WTC.
Each workstation purchased as part of the Investigation contained about 250 kg of combustible material,
mostly a mix of paper and plastic (NIST NCSTAR 1-5C). In the numerical simulations, the modeled
workstations were distributed on each floor as shown in Figure 1-1, following the architectural plan.
Outside of the core area, the combustible load was approximately 20 kg/m2 (4 lb/ft2) (NIST NCSTAR 1-
5). The core area contained elevator and HVAC shafts, stairwells, storage rooms, toilets, and various
other support facilities. It was unclear what the combustible load was in the various core areas. As a first
approximation, the carpet that was assumed to be spread over the entire floor was extended into the core
area, not necessarily because the core was carpeted but to represent whatever other combustible objects
were to be found there.
Unlike WTC 1 and 2, ceiling tiles were included some of the simulations. Each tile was assigned thermal
properties obtained from the manufacturer, and it was assumed that it became dislodged from its supports
when its surface temperature reached 400 °C.
Confidential and Predecisonal Document                                                      Title of Appendix

Figure 1-2. Photograph of a work area in the WTC, courtesy Port Authority.

1.2.3       Exterior Damage
A very important component of the fire simulations was the damage to the exterior of the building, both
from the debris impact from WTC 1 and from the heat of the fires breaking windows. There were no
attempts to predict window breakage with the model, nor was there any effort to assess the sensitivity of
the calculations to exterior damage. Exterior damage was input directly into the model from the
photographs. Gaps in the photographic evidence were “filled in” by assuming that windows broke out at a
constant rate, consistent with the assumption that the fires spread along a given face of the building at a
more or less constant rate.

1.2.4       Ignition
For the simulations of WTC 1 and 2, the ignition of furnishings was clearly a result of jet fuel, and this
effect was reproduced in the simulations using evaporating liquid droplets. For WTC 7, however, the
ignition and early course of the fires were unknown because they were assumed to have occurred in the
damaged and heavily smoke-shrouded southern portion of the building. For this reason, the simulated
fires were assumed to have originated near the southern face of the building. For each floor, a specified
fire of 2 MW (essentially a steadily burning object) was positioned near the southern face of the building.
Appendix B                                                  Confidential and Predecisional Document, 1/11/2009

This is roughly the equivalent of a small, single office workstation burning. Furnishings near the
originally specified fire were assigned thermal properties obtained from small scale experiments (NIST
NCSTAR 1-5F). Fire spread was predicted by the model as a natural consequence of surrounding objects
heating up and burning. However, for the simulation of Floor 8, a second specified fire was needed to
spread the fire to the north face of the building, where it was observed from about 3 p.m. until collapse at
5:20.

1.3          RESULTS OF THE SIMULATIONS
The aim of the fire simulations for WTC 7 was to replicate the major features of the fires given the
limited knowledge of the debris impact damage and interior contents, while exploiting as much as
possible the visible evidence contained within the few photographs and videotapes taken of WTC 7
following the collapse of the towers. The major features replicated included the rate of spread of the fires,
the duration of fire activity in a given location, the thermal insult to the structural elements, and the
overall burn time for each of the affected floors.
Note that only the window breaking times were prescribed in the fire model. The observed fire activity
gleaned from the photographs and videos was not a model input, thus one should not expect a one to one
correspondence between predicted high temperatures and observed fire activity.
Confidential and Predecisonal Document                                                        Title of Appendix

1.3.1       Floor 7
A “cubicle” fire was observed on Floor 7 at about 12:15, near the southwest corner of the building. By
3:00, the fire had progressed to about halfway along the north face, from the west side of the building.
Shortly after this time, the fire appeared to stop and was not observed to progress further. As late as 4:45,
fire was observed near the middle of the north face on Floor 7.
To simulate the fire on Floor 7, a 2 MW fire was prescribed near the southwest corner at 12:00, the
starting time of the simulation. This fire was chosen merely to initiate spread along the west face of the
building. No other fires were prescribed; the spread of the fire beyond its point of origin was predicted by
the model. Windows were removed from the calculation at times observed in the photographs and videos.
For time periods between photographs, windows were removed in a linear fashion, consistent with the
assumption that the fires spread at a nearly constant rate.
Unlike the actual fire on Floor 7, the simulated fire did not stop near the middle of the north face. There
were office walls in this vicinity, but doors were left open in the simulation and the fire eventually passed
through the middle section of the building and moved to the east face, where it finally consumed the
remainder of the furnishings. Figure 1-3 displays hourly snapshots of the upper layer temperatures
predicted by the model.
Appendix B                                                 Confidential and Predecisional Document, 1/11/2009

1:00 pm

2:00 pm

3:00 pm

4:00 pm

5:00 pm

Figure 1-3. Progression of simulated fire, Floor 7, WTC 7. Shown are upper layer temperatures, with the
color red representing severe fire activity (temperatures of about 1000 °C).
Confidential and Predecisonal Document                                                      Title of Appendix

1.3.2       Floor 8
At 12:15 when the cubicle fire was observed on Floor 7, several people were being led from Floors 7 and
8 out of the building. No fires, heavy dust, or smoke were reported at this time on Floor 8. Between 12:15
and 2:30, fire activity on Floor 8 was observed at the south face by eyewitnesses near the southwest
corner of the building. At 3:00, fire was observed on Floors 7 and 12 at the north face, but not Floor 8.
Some time later that hour, fire was observed on the north face of Floor 8, moving from west to east.
Unlike the fire on Floor 7, the fire on Floor 8 continued to move east on the north face, eventually
reaching the northeast corner and moving to the east face. By 4:45, fires were still observed on the north
face of Floor 8.
In the simulation, a 2 MW fire was prescribed near the middle of the south face at 12:30, and a second 2
MW fire was prescribed near the northwest corner at 3:00. The second fire was needed because the first
fire did not spread to the north face in preliminary simulations. The architectural drawing of Floor 8 did
not include furnishings along the west side, and it was not clear just how the actual fire spread from the
south side to the north. Thus, based on the photographic evidence, a fire was simply prescribed at the
northwest corner to initiate spread along the north face.
Appendix B                                                 Confidential and Predecisional Document, 1/11/2009

1:00 pm

2:00 pm

3:00 pm

4:00 pm

5:00 pm

Figure 1-4. Progression of simulated fire, Floor 8, WTC 7. Shown are upper layer temperatures, with the
color red representing severe fire activity (temperatures of about 1000 °C).
Confidential and Predecisonal Document                                                       Title of Appendix

1.3.3       Floors 11, 12 and 13
From 2:00 to 2:30, fire was observed on the east face of Floors 11 and 12 at the southeast corner, moving
northward. By 3:00, fire was observed on Floor 12 at the north face of the building, about one-third of the
way towards the northwest corner. It spread in both the east and west direction along the north face after
this time. Some time later, fire was observed on Floor 13, moving from east to west along the north face.
By 4:45, the fires on Floor 12 appeared to have burned out. Fire on Floor 11 was observed about one-third
of the way from east to west on the north face. Much of the north face was obscured by smoke at this
time; thus, it is difficult to be more precise about the eventual progression of the fires on Floors 11, 12
and 13.
The fire behavior on Floors 11, 12 and 13 followed similar patterns, with a slight offset in time of
between 15 to 30 min.
Appendix B                                                 Confidential and Predecisional Document, 1/11/2009

1:00 pm

2:00 pm

3:00 pm

4:00 pm

5:00 pm

Figure 1-5. Progression of simulated fire, Floor 12, WTC 7. Shown are upper layer temperatures, with the
color red representing severe fire activity (temperatures of about 1000 °C).
Confidential and Predecisonal Document                                                         Title of Appendix

1:00 pm

2:00 pm

3:00 pm

4:00 pm

5:00 pm

Figure 1-6. Progression of simulated fire, Floor 12, WTC 7, under the assumption that the ceilings tiles
withstood the heat of the fire for roughly twice as much time as the baseline cases.
Appendix B                                                 Confidential and Predecisional Document, 1/11/2009

1.4          SUMMARY OF FIRE SIMULATIONS
The simulations of the fires in WTC 7 were part of a larger analysis that included the debris impact from
WTC 1, the subsequent fires, the heating of the structural elements, and the eventual collapse.
Temperature predictions from the fire model were passed to the structural model via the Fire-Structure
Interface (NIST NCSTAR 1-5G).
Although the visual evidence for WTC 7 was not nearly as rich as for WTC 1 and 2, the fire simulations
did exploit as much as possible the few photographs showing the location of severe fire activity at various
times during the afternoon of September 11, 2001. The simulated fires followed the same general paths,
burned for about the same length of time, and as evidenced by the validation experiments, generated
comparable temperatures to those of the real fires.

```
DOCUMENT INFO
Shared By:
Categories:
Stats:
 views: 31 posted: 1/11/2009 language: English pages: 12