NRC Publications Archive (NPArC)

Archives des publications du CNRC (NPArC)

Experiments of Smoke Detector Response in Prison Facilities

Su, J. Z.; Taber, B. C.; Carpenter, D. W.

Web page / page Web

[Link]
[Link]

Access and use of this website and the material on it are subject to the Terms and Conditions set forth at
[Link]
READ THESE TERMS AND CONDITIONS CAREFULLY BEFORE USING THIS WEBSITE.

L’accès à ce site Web et l’utilisation de son contenu sont assujettis aux conditions présentées dans le site
[Link]
LISEZ CES CONDITIONS ATTENTIVEMENT AVANT D’UTILISER CE SITE WEB.

Contact us / Contactez nous: [Link]@[Link].

EXPERIMENTS OF SMOKE
DETECTOR RESPONSE IN
PRISON FACILITIES

Research Report 182

Date of Issue: November 12, 2004

Authors: Joseph Z. Su, Bruce C. Taber

and Don W. Carpenter

Published by
Institute for Research in Construction
National Research Council Canada
Ottawa, Canada
K1A 0R6
 EXPERIMENTS OF SMOKE DETECTOR RESPONSE IN PRISON FACILITIES

By

Joseph Z. Su, Bruce C. Taber, Don W. Carpenter

Fire Research Program

Institute for Research in Construction
National Research Council of Canada

ABSTRACT

Nuisance fire alarms can be a problem in Canadian correctional facilities when

inmates intentionally activate or damage in-cell smoke detectors, which are currently
required under the National Building Code of Canada. These alarms result in increased
risk to guards and inmates while the detector is out of service, time lost as guards
investigate the cause, and significant costs to examine and replace damaged detectors.

To find a solution, the Correctional Service of Canada (CSC) initiated a project

with researchers in NRC's Fire Research Program and Ken Richardson Fire
Technologies Inc. A series of full-scale tests were set up in temporarily vacated
correctional facilities in Kingston, Ontario, to determine if in-cell smoke detectors could
be moved outside of cells and still provide an equivalent level of fire protection. In
particular, the risk to inmates in the cell of fire origin should not exceed critical limits for
carbon monoxide, carbon dioxide and temperature if the detectors were located outside
the cell.

Test scenarios were representative of the fires that actually occur in cells, while
still posing a reasonable challenge for the detectors expected to respond. These
scenarios involved different fire sources (such as CSC-issue mattresses and clothing,
and newspaper) for both open-front and closed-front cells. Depending on whether the
cell had an open or closed front, the location of the fire source was varied.

In all scenarios, fire detection with smoke detectors in an exhaust duct adjacent
to a cell was equivalent to that provided by in-cell smoke detectors in both open-front
and closed-front cells. Specifically, moving smoke detectors from inside to outside
open-front cells, to either a duct or the corridor, did not affect reaction times enough to
allow critical conditions to build up in the cell where the fire originated. For closed-front
cells, only smoke detectors relocated to an exhaust duct provided an equivalent level of
fire detection. In some cases with closed-front cells, smoke detectors moved to the
corridor allowed critical conditions to build beyond acceptable levels in the cell of fire
origin.

i
EXPERIMENTS OF SMOKE DETECTOR RESPONSE IN PRISON FACILITIES

Joseph Z. Su, Bruce C. Taber, Don W. Carpenter

TABLE OF CONTENTS

ABSTRACT ...................................................................................................................... i
1.0 INTRODUCTION ....................................................................................................1
2.0 EXPERIMENTS ......................................................................................................1
2.1 Experimental Cell ...........................................................................................1
2.2 Smoke Detectors ...........................................................................................2
2.3 Fuels and Fire Scenarios ...............................................................................2
2.4 Instrumentation ..............................................................................................3
2.5 Experimental Procedure.................................................................................4
3.0 EXPERIMENTAL RESULTS...................................................................................4
4.0 CONCLUSIONS ...................................................................................................10
5.0 ACKNOWLEDGEMENTS .....................................................................................11
6.0 REFERENCE .......................................................................................................11

LIST OF FIGURES

Figure 1 Two-storey cellblock and experimental cell on the ground floor

Figure 2 Plane view of experimental set-up
Figure 3 Four smoke detectors around the existing CSC detector in the cell
Figure 4 Ten smoke detectors in the corridor under the 2-floor walkway
Figure 5 Fuel packages of cotton mattress and polyurethane foam mattress
Figure 6 Fuel packages of clothing/parka and newspaper for flaming fires
Figure 7 Measurement devices and data acquisition system
Figures 8–21 Optical density of the smoke and response of the smoke detectors in
various locations in Test 1 to Test 14.

Appendix A. Conversion chart of optical density and obscuration

Appendix B. Concentration-time profiles of carbon monoxide, carbon dioxide and oxygen
Appendix C. Fax from Siemens on response of the CSC existing detector and the
aspiration smoke detection device

ii
 EXPERIMENTS OF SMOKE DETECTOR RESPONSE IN PRISON FACILITIES

Joseph Z. Su, Bruce C. Taber, Don W. Carpenter

1.0 INTRODUCTION

The Correctional Service of Canada (CSC) has installed fire alarm systems throughout
its prison facilities, including smoke detectors in individual cells as required by Article [Link] of
the National Building Code of Canada [1]. Since inmates frequently damage the smoke
detectors inside cells, CSC is considering relocating the smoke detectors outside the cells.
CSC needs technical answers on whether this relocation increases fire detection time or risks
due to fire.

The National Research Council of Canada (NRC) has conducted a research project, for
CSC, to study the difference in response of smoke detectors installed inside and outside the
cells. The objective of the project is to determine differences in fire detection time and the
associated risk due to fires, relating to different detector locations.

This project required full-scale fire experiments to determine smoke detector response
under a variety of conditions. NRC completed a series of fire detection experiments in a prison
cell at Collins Bay Institution, with smoke detectors being installed at various locations in the
cell, corridor and exhaust duct. The results of the experiments are documented in this report.

The project also involved an independent engineering analysis of operational and

statistical data on reliability, maintenance, vandalism incidents, repairs and false alarms of
smoke detection systems in correctional facilities. The engineering analysis and fire risk
assessment was conducted and reported separately by Ken Richardson Fire Technologies Inc.

2.0 EXPERIMENTS

On January 28-30, 2002, NRC conducted 14 full-scale experiments at the Collins Bay
Institution. The experiments are intended to evaluate the differences in response time for
smoke detectors located in a cell compared to those located outside the cell envelope as part of
an evaluation of risk to all occupants.

2.1 Experimental Cell

The experimental cellblock has 2 levels of cells with open-bar doors and a 2-storey high
common hallway, as shown in Figure 1. A cell on the ground floor of the cellblock was used for
fire detection experiments. The cell was 3.37 m long, 1.66 m wide, and 2.44 m high. The cell
open-bar front was 1.21 m wide and 2.34 m high, with a 10 cm lintel. The top 30 cm of the open
bars was covered with plywood to create additional lintel height. In about half of the
experiments, the open-bar front was completely covered using plywood to simulate a solid door
cell.

At the back of the cell, there were a toilet and a washbasin. An existing sidewall
sprinkler head (rated at 71°C or 160°F) and a duct opening (15 cm diameter) to the exhaust
system were located on the back wall at a 2.1 m height.

1
2.2 Smoke Detectors

The cell had an existing CSC smoke detector on the ceiling, which was set to alarm at
an obscuration of 3% per 304 mm (an optical density of 0.0435 per meter). This is the detector
that CSC is considering moving. CSC also had a portable aspiration smoke detection device
with a sampling tube installed in the cell, which displayed obscuration readout. This device had
4 alarm levels: alert at 0.025%/304 mm (3.6x10 -4/m), action at 0.0437%/304 mm (6.2x10 -4/m),
pre-fire alarm at 0.0625%/304 mm (8.9x10 -4/m), fire alarm at 3.0%/304 mm (0.0435/m).

In order to achieve experiment objectives in the determination of differences in fire

detection time with different detector locations, multiple smoke detectors needed to be installed
at various locations inside and outside the cell. This could not be done with the existing CSC
fire alarm system. It was decided that CSC officials would arrange to monitor their existing fire
alarm system and the aspiration detection device and that NRC would install a stand-alone
detector system for the experiments.

A total of 7 photoelectric and 7 ionization smoke detectors (all powered by batteries)

were installed inside and outside the experimental cell. Figure 2 shows a plan view of the
locations of all smoke detectors installed. The existing CSC smoke detector in the cell (at mid-
width, 2 m from the back wall) was protected by a metal cage (appears as a square in Figure 2).
Two ionization detectors and 2 photoelectric detectors were installed in the cell around the
existing CSC detector at the ceiling, as shown in Figure 3. Each of these 4 detectors had its
edge 100mm away from the cage of the existing detector and centered on each side of the
cage. Five ionization detectors and 5 photoelectric detectors were installed in the corridor
outside the cell (under the 2nd-floor walkway), as shown in Figures 1 and 4. An ionization
detector and a photoelectric detector formed a pair. One pair of detectors were parallel with and
centered on the cell door, 38 cm from the wall. Four pairs of detectors were lined up
perpendicular with, and centered on, the four adjacent pillars between the cells. Each pair had
one detector 25 cm from the pillar and the other detector 49 cm from the pillar. For each pair,
the two detectors were spaced 100 mm from edge to edge.

There was no smoke detector installed in the exhaust duct from the cell. However,
optical density in the duct was measured, providing data for determination of would-be
activation time should a detector be located in the duct.

Two sprinkler heads with a temperature rating of 71°C (160°F) were mounted on the
back wall beside the existing sprinkler head for possible "dry" operation during experiments.
The existing automatic sprinkler system was de-actuated during the experiments.

2.3 Fuels and Fire Scenarios

Small and slow-growing fires (both flaming and smouldering) were used as test fires
since these represent the most challenging situation for smoke detectors and the worst-case
scenario as far as fire detection was concerned. Experimental fuel packages were prepared
and pre-tested by NRC using cell content materials provided by CSC.

A cotton mattress and a polyurethane foam mattress were cut into 20 cm x 20 cm

square pieces as fuel packages for smouldering fires. Each cotton mattress piece weighed

2
approximately 330 g and each foam mattress piece weighed approximately 290 g. As shown in
Figure 5, smouldering fires were produced by putting the mattress fuel package on an
electrically-powered barbecue lighter, which lay on ceramic fibre batt insulation on the floor. A
perforated metal bucket was used to cover the fuel and ignitor. The holes around the metal
bucket controlled the amount of air available for combustion. Generated smoke came out from
the holes on the top of the metal bucket.

Pieces of clothing (parka) and newspaper were used for flaming fires, as shown in
Figure 6. Clothing fuel packages were made using pieces of parka with denim material on top.
One arrangement of the clothing packages was a strip of parka with the denim on top (12.7 cm
long, 61 cm wide, 81 g). Another arrangement was 6 layers of parka and 6 layers of the denim
stacked alternately over each other (20 cm long, 15 cm wide, 195 g). Flaming clothing fires
were produced by pilot ignition using a torch. Ten full sheets of newspaper (weighed 190 g)
were folded 8 times to a size of 15 cm x 15 cm. The newspaper package was put on a pre-
heated barbecue lighter and ignited using a torch.

The fuel packages used in the experiments represent the most common materials first
ignited in fire incidents in correctional occupancies. In order to provide the greatest challenges
to the operation of the detectors located outside the cell envelope, fire locations were varied
depending on whether the cell had an open or closed front, as shown in Figure 2. For
experiments with the open front, the fire location was near the back wall since that was the
greatest distance from the corridor detectors. For experiments with the closed front, the fire
location was near the solid front door since that was the greatest distance from the exhaust
duct. One experiment was also conducted using sweet grass as fuel in the open-bar cell.

2.4 Instrumentation

Various devices were used to measure smoke detector response, smoke obscuration
(optical density), gas concentrations, temperatures and automatic sprinkler response, as shown
in Figures 1–4 and 7. Experimental data was collected using a data acquisition system located
in an adjacent cell. Data also included airflow in the exhaust duct of the cell and video records.

The response times of all detectors were determined by measuring the current draw
through the batteries. Electric resistance of the sprinkler heads was also monitored to
determine whether the sprinklers would operate.

Smoke optical density was measured as a function of time at 3 different locations in

each experiment. The three optical density meters were of the aspiration type, drawing air
samples from the ceiling near Detectors #12– #15 in the cell; from the location near Detectors
#10– #11 in the corridor; and from the exhaust duct of the cell.

Temperatures in the cell were measured using a thermocouple tree. The thermocouple
tree was installed near the centre of the cell beside Detector #15, with 5 thermocouples
distributed at heights of 0.6, 1.5, 1.8, 2.1 and 2.4 m above the floor.

CO, CO2 and O2 concentrations in the cell were measured using 4 gas analyzers. The
gas sampling ports were mounted on the thermocouple tree at heights of 1.5 and 2.4 m above
the floor in the cell.

3
2.5 Experimental Procedure

All experiments were run with the normal building/cell ventilation system operational.
The experimental procedure was as follows:

1. Verification of experiment layout, instrumentation and data acquisition;

2. t = 0 (time zero), starting data acquisition system;
3. t = 1 minute, igniting fuel;
4. Observation of smoke movement;
5. Ending the experiment.

The duration of each experiment was sufficient to enable determination of differences in

detector response between in-cell and outside-cell units. The following represent the principles
for experiment termination:

• For the open front experiments, response of in-cell detectors and corridor detectors directly
in front of cell door (Detectors #10– #15) plus an additional 5 min (or more) to determine if
detectors located at junctions between cells would operate;
• For the closed front experiments, response of in-cell detectors (Detectors #12– #15) and the
optical density in duct exceeding the detector activation value; or
• A ceiling temperature of 50°C.

At the end of each experiment, NRC staff manually extinguished the fire by turning off
the power supply to the barbecue lighter and carefully dropping the fuel package into a bucket
of water or by using fire blankets.

CSC officials arranged with Siemens Building Technologies to monitor the existing fire
alarm system and the aspiration smoke detection device during the experiments.

3.0 EXPERIMENTAL RESULTS

Table 1 shows the matrix of the fire detection experiments. The airflow in the exhaust
duct was measured for both the open and closed front cells using a portable manometer. The
air velocity at the centre of the exhaust duct was 3.4 m/s for the open-bar cell (measured during
Test 4) and 2.7 m/s for the closed front cell (measured during Test 11).

Table 2 shows the maximum changes of temperature and gas concentrations in the cell
during each experiment. The initial temperature in the cell was 24–26°C for all experiments.
Since the test fires were small and grew slowly (in order to challenge the smoke detectors), the
temperature rise in the cell at the ceiling was less than 4°C for all smouldering experiments and
less than 21°C for all flaming experiments. In other words, the ceiling temperature in the cell
was less than 30°C for all smouldering experiments and less than 46°C for all flaming
experiments. The dry sprinkler heads did not actuate during any experiments.

Table 2 also shows activated smoke detectors (with grey shading) during the fire
detection experiments and smoke optical densities (ODf) at the times of activation of Detectors
#12– #15 and Detectors #10– #11. In most cases when the cell had a solid door (closed), the
smoke detectors outside the cell did not detect the fires inside the cell. When the cell had an
open-bar door, the outside detectors detected the inside fires in most cases.

4
 Figures 8–21 show optical density (ODf) of the smoke and response of the smoke
detectors in various locations (Note that ignition was at 60 s). The response times of the smoke
detectors from ignition are shown in Table 3. For the flaming fires, the ionization detectors were
almost always activated earlier than the photoelectric detectors at the same locations relative to
the fires. For the smouldering fires, only in a little more than half of the cases, the photoelectric
detectors were activated earlier than the ionization detectors at the same locations relative to
the fires. Foam mattress smouldering took the longest time to detect under similar conditions.

The air movement in the hallway during the experiments, as indicated in Figure 2, had
an impact on the response of the hallway smoke detectors. Because the air moved toward the
east in the hallway, the detectors in the east side of the experimental cell generally activated
earlier than those in the west side of the cell.

For the open-bar scenarios, the corridor detectors detected the test fires inside the cell
within reasonable times. For the flaming fires in the open-bar cell, the first smoke alarm in the
corridor occurred in less than 1 minute after the first smoke alarm in the cell. For the
smouldering fires in the open-bar cell, the first smoke alarm in the corridor occurred 1 to
4 minutes after the first smoke alarm in the cell. (All these are less than 60% longer than the
in-cell detector response.)

From the experimental placement of the smoke detectors, 4 different placements of

corridor detectors can be extracted for possible applications in CSC facilities. Placement A
includes all the corridor detectors installed as shown in Figure 2. Placement B includes the
corridor detectors installed in front of every pillar. Placement C includes only Detectors #6, #7,
#16 and #17 installed in front of every other pillar. Placement D includes only Detectors #8, #9,
#18 and #19 installed in front of every other pillar. Table 3 shows the time delay of the first
smoke alarm in the corridor from the first smoke alarm in the cell for the 4 different corridor
detector placements. There are no significant differences in the time delay for Placements A
and B, indicating that the corridor detectors could be centered either on every cell or on every
pillar with similar overall performance. Placements C and D are two arrangements for detector
spacing at every other pillar. Favoured by the air movement in the corridor, Placement C
responded to the in-cell fires with a similar time delay as for Placements A and B. Placement D
would be the worse case for every-other-pillar spacing, and the time delay for this case should
be used to determine whether detector spacing at every other pillar is acceptable.

No corridor detectors responded to the fires inside the closed cell in Tests 7, 8 and 12.
Only a few corridor detectors initiated an alarm to the fires inside the closed cell in Tests 9, 10
and 11, with the earliest response time of 2 to 9 minutes (about 5 times) longer than the in-cell
detector response.

Since the corridor smoke detectors could not reliably detect the test fires inside the cell
when the cell had a solid door (closed), it is necessary to determine whether a duct detector is a
viable option. Comparing the optical density profile in the duct with that in the cell (as shown in
Figures 8–21) will allow determination of the time delay to reach a specific optical density in the
duct. Table 4 shows the time delays of optical density in the duct at 7 different smoke
obscuration levels, compared to the times when optical density in the cell reached the same
levels. The optical density values of 0.007, 0.014, 0.029, 0.044, 0.058, 0.073 and 0.088 per
meter in the table are equivalent to the smoke obscuration levels of 0.5, 1, 2, 3, 4, 5 and 6% per
foot, respectively. (Appendix A shows a conversion chart of optical density and obscuration.)
The value of the obscuration per foot represents the percentage of projected light that cannot
reach the distance of 1 foot. Under various experimental conditions, the delay times were less

5
than 1 minute in 60% of the cases and less than 2 minutes in 90% of the cases. At a 0.044/m
optical density (3% per foot obscuration), the delay times in the duct were less than 2 minutes in
all experiments. The delay time in the duct was usually longer for a higher optical density. The
longest delay time was 6.5 minutes at the 0.088/m optical density (6% per foot obscuration) in
Test 12. Should a smoke detector be installed in the duct, its response time to fires would
depend not only on the fire conditions but also on its sensitivity setting (the lower the optical
density, the higher the sensitivity).

The CO and CO2 concentrations listed in Table 2 are the maximum values in each
experiment. For all flaming fires, it took more than 4 minutes for CO and CO2 to reach these
maximum concentrations. For all smouldering fires, it took more than 5 minutes for CO and CO2
to reach these maximum concentrations. In most experiments, the time to reach the maximum
CO and CO2 concentrations was well behind the activation of the majority of the detectors.
(Complete concentration-time profiles are included in Appendix B.) The flaming fires produced
more CO2 while the smouldering fires produced more CO; the fires in the closed cell produced
higher CO and CO2 concentrations. Test 9 (clothing/parka flaming, closed cell) produced the
highest CO2 concentration of 0.6% and the lowest O2 concentration of 20.3% (as well as the
highest ceiling temperature of 46°C). Test 11 (cotton mattress smouldering, closed cell)
produced the highest CO concentration of 310-340 ppm. These CO and CO2 concentrations
produced under the experiment conditions are much lower than the limiting values that are
immediately dangerous to life or health (IDLH). The IDLH value is 1500 ppm for CO and 5% for
CO2 [2].

The results of response from the existing fire alarm system and the aspiration smoke
detection device monitored by CSC and Siemens are included in Appendix C. Except for
Tests 2 and 4, these results are consistent with NRC's measurements.

Table 1 – Experiment Matrix for Smoke Detector Response

Test
Burning Mode Fuel Package
Open Bar Front Solid Door Front

Clothing/Parka 1, 2, 6 7*, 8, 9
Flaming
Newspaper 3 10

Cotton Mattress 4 11
Smouldering
Foam Mattress 5, 13 12

Sweet Grass 14

* Test 7 - Fire located at the back of the cell.

6
 Table 2 – Detector Response and Maximum Changes of Temperature and Gas Concentrations (Jan. 2002)

Test 1 2 3 4 5 6 7* 8 9 10 11 12 13 14
Cell Door
Open Open Open Open Open Open Closed Closed Closed Closed Closed Closed Open Open
Clothing Clothing News- Cotton Foam Clothing Clothing Clothing Clothing News- Cotton Foam Foam Sweet
Fuel Parka Parka paper Mattress Mattress Parka Parka Parka Parka paper Mattress Mattress Mattress Grass
(strip) (stack) (stack) (strip) (strip) (stack)
Burning Flaming Flaming Flaming Smould- Smould- Flaming Flaming Flaming Flaming Flaming Smould- Smould- Smould- Smould-
Mode ering ering ering ering ering ering
Temperature 29.6 39.7 41.7 26.8 26.0 37.8 33.0 36.4 45.3 36.0 27.7 27.0 25.1 28.6
Max. (°C)
Drop Rise (ppm)

ceiling 20 15 90 240 150 20 40 65 70 160 340 190 150 15

O2% CO2% CO

1.5-m 30 10 50 150 70 20 50 60 60 130 310 160 70 10

ceiling 0.06 0.18 0.18 0.05 0.02 0.15 0.21 0.32 0.55 0.25 0.08 0.07 0.03 0.03
1.5-m 0.04 0.12 0.09 0.04 0.01 0.11 0.21 0.31 0.54 0.25 0.09 0.07 0.01 0.01
ceiling 0.07 0.20 0.19 0.05 0 0.15 0.20 0.35 0.61 0.25 0.10 0.07 0 0
1.5-m 0.03 0.10 0.18 0.03 0 0.10 0.20 0.32 0.57 0.25 0.10 0.07 0 0
6p
7i
Detector Response ODf (1/m)

8p
9i
10 p 0.083 0.033 0.074 0.08 0.06 0.030 0.048
11 i 0.0025 0.025 0.010 0.070 0.01 0.03 0.010 0.025 0.048
12 p 0.095 0.040 0.050 0.03 0.104 0.046 0.07 0.026 0.090 0.062 0.032
13 i 0.001 0.048 0.018 0.050 0.08 0.018 0.005 0.002 0.02 0.014 0.055 0.091 0.090 0.004
14 p 0.10 0.050 0.020 0.048 0.11 0.046 0.08 0.020 0.060 0.060 0.049
15 i 0.001 0.045 0.018 0.098 0.03 0.015 0.001 0.003 0.02 0.016 0.045 0.016 0.034 0.004
16p
17 i
18 p
19 i

Notes: (1) no additional lintel for Test 1, (2) initial temperature 24–26°C, (3) i for ionization, p fo r photoelectric, grey shading for activated detectors,
(4) smoke optical density, ODf (1/m), only measured near Detectors #10, 11, 12, 13, 14 and 15, (5)* Test 7 - Fire located at the back of the cell.

7
 Table 3 – Response Time of Detectors and Time Delay of Corridor Detectors Relative to Cell Detectors

Test 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Cell Open Open Open Open Open Open Closed Closed Closed Closed Closed Closed Open Open
Clothing Clothing News- Cotton Foam Clothing Clothing Clothing Clothing News- Cotton Foam Foam Sweet
Fuel Parka Parka paper Mattress Mattress Parka Parka Parka Parka paper Mattress Mattress Mattress Grass
(strip) (stack) (stack) (strip) (strip) (stack)
Burning Flaming Flaming Flaming Smould- Smould- Flaming Flaming Flaming Flaming Flaming Smould- Smould- Smould- Smould-
Mode ering ering ering ering ering ering
6p
7i 578
Detector Response Time (s)

8p 172 355 400 660 229 230 748

9i 153 102 448 680 173 253 885
10 p 162 197 266 640 209 224 762 619
11 i 286 62 52 252 630 72 147 303 696 611
12 p 150 206 222 397 194 529 58 48 200 326 417
13 i 354 43 39 216 551 51 83 32 20 36 214 420 795 63
14 p 153 223 200 414 200 531 64 45 209 318 544
15 i 91 34 38 177 404 47 39 49 20 42 143 230 427 62
16p 127 189 260 691 171 240 771 611
17 i 144 54 53 253 717 74 164 234 620
18 p 225 710 342
19 i 751 102 85 535 728 145
A 53 20 14 75 226* 25 127 198 553 194
Delay (s)
Time

B 53 20 15 76 263 27 144 198 571* 194

C 53 20 15 76 294 27 144 198 571* 194
D 660 68 47 200* 263 98 172* 331
Notes:
(1) No additional lintel for Test 1;
(2) Initial temperature of 24–26°C;
(3) Grey shading for activated detectors, i for ionization detector, p for photoelectric detector;
(4) Time Delay = time of the first corridor detector activation minus time of the first in-cell detector activation (* to ensure that the detectors
involved are of the same type, this value is the time of the 1st corridor detector activation minus time of the 2nd in-cell detector activation) for
• Corridor Detector Placement A: containing Detectors #6, #7, #8, #9, #10, #11, #16, #17, #18, #19
• Corridor Detector Placement B: containing Detectors #6, #7, #8, #9, #16, #17, #18, #19 (in front of every pillar)
• Corridor Detector Placement C: containing Detectors #6, #7, #16, #17 (in front of every other pillar)
• Corridor Detector Placement D: containing Detectors #8, #9, #18, #19 (in front of every other pillar)

8
 Table 4 – Time Delay (in seconds) of Duct Optical Density Relative to Cell Optical Density

Test 1 2 3 4 5 6 7 8 9 10 11 12 13
Cell
Open Open Open Open Open Open Closed Closed Closed Closed Closed Closed Open
Clothing Clothing News- Cotton Foam Clothing Clothing Clothing Clothing News- Cotton Foam Foam
Fuel Parka Parka paper Mattress Mattress Parka Parka Parka Parka paper Mattress Mattress Mattress
(strip) (stack) (stack) (strip) (strip) (stack)
Burning Flaming Flaming Flaming Smould- Smould- Flaming Flaming Flaming Flaming Flaming Smould- Smould- Smould-
Mode ering ering ering ering ering
0.007/m 110 12 10 29 -17 10 20 34 30 25 34 34 40
Optical Density

0.014/m 11 12 20 22 20 39 100 32 32 46 23 40
0.029/m 23 39 -32 42 35 105 34 45 51 70 103
0.044/m 55 -26 75 32 110 38 110 54 112 105
0.058/m 64 10 98 70 45 115 144 136 120
0.073/m 28 24 65 43 117 190 240
0.088/m 43 105 46 113 200 395

9
4.0 CONCLUSIONS

NRC completed a series of fire detection experiments on smoke detector response in

the prison cell at the Collins Bay Institution to study the difference in response of smoke
detectors installed inside and outside the cells. NRC installed stand-alone detector arrays with
14 photoelectric and ionization smoke detectors (powered by batteries) at various locations
inside and outside the experimental cell. All experiments were run with the normal building/cell
ventilation system operational.

Small and slow-growing fires (both flaming and smouldering) were used as test fires
since they would represent the most challenging situation for smoke detectors and the worst-
case scenario as far as fire detection was concerned. Experimental fuel packages were
prepared using the cell content materials, including cotton mattress, polyurethane foam
mattress, clothing/parka and newspaper, which represented the most common materials first
ignited in fire incidents in correctional occupancies.

Since the test fires were small, the ceiling temperature in the cell never exceeded 46°C
and, therefore, the automatic sprinklers were never actuated. The CO and CO2 concentrations
produced under the experiment conditions were much lower than the limiting values that are
normally considered to represent an immediate danger to the life or health of a healthy adult.

At the same locations relative to the fires, the ionization detectors always responded to
the flaming fires earlier than the photoelectric detectors while the response of the photoelectric
detectors to the smouldering fires was not always earlier than the ionization detectors. Foam
mattress smouldering took the longest time to detect.

For the open-bar scenarios, the corridor detectors detected the test fires inside the cell
with the earliest response time less than 60% longer than the in-cell detector response. With
the corridor detectors installed in front of every pillar, fires inside the open-bar cell are likely to
be detected. However, should the corridor detectors be spaced at every other pillar, their
response time to fires inside the cell could be double or triple that of the in-cell detectors.

In most cases where the cell had a solid door (closed), the smoke detectors outside the
cell could not detect the fires inside the cell. In cases where the corridor detectors initiated an
alarm to the fires inside the closed cell, the earliest response time was at least 5 times longer
than the in-cell detector response.

The optical density measurement in the duct indicated that a duct detector works for the
closed cell, provided that the duct detector be set at an appropriate sensitivity. In all
experiments, the time delay for the optical density to reach 0.044/m (3% per foot obscuration) in
the duct was less than 2 minutes from the time when the optical density in the cell reached the
same level. This delay time would be longer for a higher optical density. Should a smoke
detector be installed in the duct, its response time to fires would depend not only on the fire
conditions but also on its sensitivity setting (the lower the optical density, the higher the
sensitivity, and the shorter the detection time).

With the results from these experiments and the analysis of statistical data on reliability,
maintenance and vandalism incidents of smoke detection systems in correctional facilities, CSC
can determine fire risk associated with different detector locations and propose a reliable, cost-
effective alternative to in-cell smoke detectors.

10
5.0 ACKNOWLEDGEMENTS

This study was conducted as a joint research project between the National Research
Council of Canada and the Correctional Service of Canada. NRC/IRC recognizes CSC's
contribution in co-funding this project and providing the logistic support. In particular, the
contribution of Mr. Randy Gaw, Chief Operational Fire Safety of CSC, in providing technical
input and guidance and necessary assistance to NRC in designing and conducting the
experiments is gratefully acknowledged.

6.0 REFERENCE

1. Canadian Commission on Building and Fire Codes, "National Building Code of Canada,"
National Research Council of Canada, Ottawa, 1995.
2. Occupational Safety and Health Administration, "OSHA Regulated Hazardous Substances,"
Noyes Data Corp., Park Ridge, 1990.

11
Figure 1. Two-storey cellblock and experimental cell on the ground floor (Contract B4131).
 service corridor

OD
exhaust duct

SS LEGEND

fire location for open cell

P photoelectric smoke detector

I ionization smoke detector

existing smoke detector
OD
GS
TC light fixture S sprinkler head
15
I
12 P P 14 OD optical density measurements
I (at ceiling in cell and walkway)
13
GS gas sampling at ceiling and 5'
fire location for closed cell (CO, CO2, O2)

TC thermocouples (2', 5', 6', 7', ceiling)

pillar pillar pillar pillar

N
I 7 P 8 I 17 P 18
I P ↑
P 6 I 9 P 16 I 19
11 10 W ←→ E
edge of level 2 walkway OD air movement ↓
S
Figure 2. Plan view of experimental set-up (Contract B4131).
Figure 3. Four smoke detectors around the existing CSC detector in the cell (Contract B4131).

Figure 4. Ten smoke detectors in the corridor under the 2-floor walkway (Contract B4131).
Figure 5. Fuel packages of cotton mattress and polyurethane foam mattress on an electrical barbecue lighter covered by a
perforated metal bucket for smouldering fires (Contract B4131).
Figure 6. Fuel packages of clothing/parka and newspaper for flaming fires (Contract B4131).
Figure 7. Measurement devices and data acquisition system (left - duct optical density measurement in the service corridor;
middle - detector response, smoke density, gas and temperature measurements in the experiment cell; right - data acquisition
system in an adjacent cell. Contract B4131).
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10

1100 1100
■ alarm activated
1000 ❑ no alram 1000

900 900

800 800

700 700
Time [s]

Time [s]
600 600

500 500

400 400

300 300

200 200
corridor
100 cell 100
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10
-1
Detector Optical Density ODf (m )
Figure 8. Test 1 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.05 0.00 0.05 0.10 0.15 0.20

800 800

■ alarm activated
❑ no alram
700 700

600 600

500 500
Time [s]

Time [s]
400 400

300 300

200 200

corridor
100 100
cell
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.05 0.00 0.05 0.10 0.15 0.20
-1
Detector Optical Density ODf (m )
Figure 9. Test 2 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08

700 700
■ alarm activated
❑ no alram
600 600

500 500

400 400
Time [s]

Time [s]
300 300

200 200

100 corridor 100

cell
duct
0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08
-1
Detector Optical Density ODf (m )
Figure 10. Test 3 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
1000 1000
■ alarm activated
900
❑ no alram 900

800 800

700 700

600 600
Time [s]

Time [s]
500 500

400 400

300 300

200 200
corridor
100 cell 100
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
-1
Detector Optical Density ODf (m )
Figure 11. Test 4 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12

1000 1000
■ alarm activated
❑ no alram
900 900

800 800

700 700

600 600
Time [s]

Time [s]
500 500

400 400

300 300

200 200
corridor
100 cell 100
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12
-1
Detector Optical Density ODf (m )
Figure 12. Test 5 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18
600 600

■ alarm activated
❑ no alram

500 500

400 400
Time [s]

Time [s]
300 300

200 200

100 corridor 100

cell
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18
-1
Detector Optical Density ODf (m )
Figure 13. Test 6 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10

800 800
■ alarm activated
❑ no alram
700 700

600 600

500 500
Time [s]

Time [s]
400 400

300 300

200 200

corridor vs time
100 100
cell vs time
duct vs time

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10
-1
Detector Optical Density ODf (m )
Figure 14. Test 7 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08

700 700
■ alarm activated
❑ no alram
600 600

500 500

400 400
Time [s]

Time [s]
300 300

200 200

100 corridor 100

cell
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08
-1
Detector Optical Density ODf (m )
Figure 15. Test 8 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6

500 500
■ alarm activated
❑ no alram

400 400

300 300
Time [s]

Time [s]
200 200

100 100
corridor
cell
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6
-1
Detector Optical Density ODf (m )
Figure 16. Test 9 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
600 600
■ alarm activated
❑ no alram

500 500

400 400
Time [s]

Time [s]
300 300

200 200

100 corridor 100

cell
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
-1
Detector Optical Density ODf (m )
Figure 17. Test 10 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.05 0.00 0.05 0.10 0.15 0.20 0.25

1000 1000
■ alarm activated
❑ no alram
900 900

800 800

700 700

600 600
Time [s]

Time [s]
500 500

400 400

300 300

200 200

corridor
100 cell 100
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.05 0.00 0.05 0.10 0.15 0.20 0.25
-1
Detector Optical Density ODf (m )
Figure 18. Test 11 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
1000 1000
■ alarm activated
❑ no alram
900 900

800 800

700 700

600 600
Time [s]

Time [s]
500 500

400 400

300 300

200 200
corridor
100 cell 100
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16
-1
Detector Optical Density ODf (m )
Figure 19. Test 12 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12

1100 1100
■ alarm activated
❑ no alram
1000 1000

900 900

800 800

700 700
Time [s]

Time [s]
600 600

500 500

400 400

300 300

200 200
corridor
cell
100 100
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.02 0.00 0.02 0.04 0.06 0.08 0.10 0.12
-1
Detector Optical Density ODf (m )
Figure 20. Test 13 - Detector Response (Contract B4131)
 -1
Detector Optical Density ODf (m )

6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10

300 300
■ alarm activated
❑ no alram

200 200
Time [s]

Time [s]
100 100

corridor
cell
duct

0 0
6 7 8 9 10 11 12 13 14 15 16 17 18 19 -0.04 -0.02 0.00 0.02 0.04 0.06 0.08 0.10
-1
Detector Optical Density ODf (m )
Figure 21. Test 14 - Detector Response (Contract B4131)
 Appendix A. Conversion Chart of Optical Density and Obscuration

15

14

13

12

11

10
Obscuration, Ou (% / 304 mm)

8 Ou ODf
-4
0.0250 3.5719x10
7 0.0437 6.2443x10-4
-4
0.0625 8.9315x10
-3
0.1400 2.0014x10
6 1.0500 0.0151
2.0000 0.0289
5 2.4000 0.0347
2.5000 0.0362
2.6000 0.0376
4 2.7000 0.0391
3.0000 0.0435
3.1000 0.0450
3 4.8000 0.0703
5.0000 0.0733
2

0
0.00 0.05 0.10 0.15 0.20 0.25

Optical Density, ODf (1/m)

 Appendix B

Concentration-time profiles of carbon monoxide, carbon dioxide and oxygen

 B4131 Test 1 - CO

0.004 time vs CO% ceiling

time vs CO% mid

0.002
CO [%]

0.000

-0.002

-0.004

0 200 400 600 800 1000 1200

Time [s]
 B4131 Test 1 - CO2

0.10

time vs CO2% ceiling

time vs CO2% mid
0.08

0.06
CO2 [%]

0.04

0.02

0.00
0 200 400 600 800 1000 1200
Time [s]
 B4131 Test 1 - O2

22.0

time vs O2% ceiling

time vs O2% mid

21.5
O2 [%]

21.0

20.5

20.0
0 200 400 600 800 1000 1200
Time [s]
 B4131 Test 2 - CO

0.004 time vs CO% ceiling

time vs CO% mid

0.002
CO [%]

0.000

-0.002

-0.004

0 200 400 600 800

Time [s]
 B4131 Test 2 - CO2

0.22

0.20 time vs CO2% ceiling

time vs CO2% mid
0.18

0.16

0.14
CO2 [%]

0.12

0.10

0.08

0.06

0.04

0.02

0.00
0 200 400 600 800
Time [s]
 B4131 Test 2 - O2

22.0

time vs O2% ceiling

time vs O2% mid
21.5
O2 [%]

21.0

20.5

20.0
0 200 400 600 800
Time [s]
 B4131 Test 3 - CO

0.010

time vs CO% ceiling

time vs CO% mid
0.008

0.006
CO [%]

0.004

0.002

0.000

-0.002
0 100 200 300 400 500 600 700
Time [s]
 B4131 Test 3 - CO2

0.25

time vs CO2% ceiling

time vs CO2% mid
0.20

0.15
CO2 [%]

0.10

0.05

0.00
0 100 200 300 400 500 600 700
Time [s]
 B4131 Test 3 - O2

22.0

21.5 time vs O2% ceiling

time vs O2% mid
O2 [%]

21.0

20.5

20.0
0 100 200 300 400 500 600 700
Time [s]
 B4131 Test 4 - CO

0.025

0.020

0.015
CO [%]

0.010

0.005

0.000
time vs CO% ceiling
time vs CO% mid
-0.005
0 200 400 600 800 1000
Time [s]
 B4131 Test 4 - CO2

0.08
time vs CO2% ceiling
time vs CO2% mid
0.07

0.06

0.05
CO2 [%]

0.04

0.03

0.02

0.01
0 200 400 600 800 1000
Time [s]
 B4131 Test 4 - O2

22.0

time vs O2% ceiling

time vs O2% mid
21.5
O2 [%]

21.0

20.5

20.0
0 200 400 600 800 1000
Time [s]
 B4131 Test 5 - CO

0.020

time vs CO% ceiling

time vs CO% mid
0.015

0.010
CO [%]

0.005

0.000

-0.005
0 100 200 300 400 500 600 700
Time [s]
 B4131 Test 5 - CO2

0.06

time vs CO2% ceiling

0.05 time vs CO2% mid

0.04
CO2 [%]

0.03

0.02

0.01

0.00
0 100 200 300 400 500 600 700
Time [s]
 B4131 Test 5 - O2

22.0

time vs O2% ceiling

time vs O2% mid
21.5
O2 [%]

21.0

20.5

20.0
0 100 200 300 400 500 600 700
Time [s]
 B4131 Test 6 - CO

0.010

time vs CO% ceiling

time vs CO% mid
0.008

0.006
CO [%]

0.004

0.002

0.000

-0.002
0 100 200 300 400 500 600
Time [s]
 B4131 Test 6 - CO2

0.20
time vs CO2% ceiling
0.18 time vs CO2% mid

0.16

0.14

0.12
CO2 [%]

0.10

0.08

0.06

0.04

0.02

0.00
0 100 200 300 400 500 600
Time [s]
 B4131 Test 6 - O2

22.0

21.5 time vs O2% ceiling

time vs O2% mid
O2 [%]

21.0

20.5

20.0
0 100 200 300 400 500 600
Time [s]
 B4131 Test 7 - CO

0.010

time vs CO% ceiling

time vs CO% mid
0.008

0.006
CO [%]

0.004

0.002

0.000

-0.002
0 200 400 600 800
Time [s]
 B4131 Test 7 - CO2

0.25

time vs CO2% ceiling

time vs CO2% mid
0.20

0.15
CO2 [%]

0.10

0.05

0.00
0 200 400 600 800
Time [s]
 B4131 Test 7 - O2

22.0

time vs O2% ceiling

time vs O2% mid
21.5
O2 [%]

21.0

20.5

20.0
0 200 400 600 800
Time [s]
 B4131 Test 8 - CO

0.010

time vs CO% ceiling

time vs CO% mid
0.008

0.006
CO [%]

0.004

0.002

0.000

-0.002
0 100 200 300 400 500 600 700
Time [s]
 B4131 Test 8 - CO2

0.4

time vs CO2% ceiling

time vs CO2% mid
0.3

0.2
CO2 [%]

0.1

0.0

0 100 200 300 400 500 600 700

Time [s]
 B4131 Test 8 - O2

22.0

time vs O2% ceiling

time vs O2% mid

21.5
O2 [%]

21.0

20.5

20.0
0 100 200 300 400 500 600 700
Time [s]
 B4131 Test 9 - CO

0.008

time vs CO% ceiling

time vs CO% mid
0.006

0.004
CO [%]

0.002

0.000

-0.002
0 100 200 300 400 500
Time [s]
 B4131 Test 9 - CO2

0.7

0.6 time vs CO2% ceiling

time vs CO2% mid
0.5

0.4
CO2 [%]

0.3

0.2

0.1

0.0

0 100 200 300 400 500

Time [s]
 B4131 Test 9 - O2

21.2

time vs O2% ceiling

time vs O2% mid
21.0

20.8
O2 [%]

20.6

20.4

20.2
0 100 200 300 400 500
Time [s]
 B4131 Test 10 - CO

0.020

time vs CO% ceiling

time vs CO% mid
0.015

0.010
CO [%]

0.005

0.000

-0.005
0 100 200 300 400 500 600
Time [s]
 B4131 Test 10 - CO2

0.30

time vs CO2% ceiling

0.25 time vs CO2% mid

0.20
CO2 [%]

0.15

0.10

0.05

0.00
0 100 200 300 400 500 600
Time [s]
 B4131 Test 10 - O2

22.0

time vs O2% ceiling

time vs O2% mid
21.5
O2 [%]

21.0

20.5

20.0
0 100 200 300 400 500 600
Time [s]
 B4131 Test 11 - CO

0.04

time vs CO% ceiling

time vs CO% mid
0.03

0.02
CO [%]

0.01

0.00

-0.01
0 200 400 600 800 1000
Time [s]
 B4131 Test 11 - CO2

0.12

time vs CO2% ceiling

0.10 time vs CO2% mid

0.08
CO2 [%]

0.06

0.04

0.02

0.00
0 200 400 600 800 1000
Time [s]
 B4131 Test 11 - O2

22.0

time vs O2% ceiling

time vs O2% mid
21.5
O2 [%]

21.0

20.5

20.0
0 200 400 600 800 1000
Time [s]
 B4131 Test 12 - CO

0.025

time vs CO% ceiling

time vs CO% mid
0.020

0.015
CO [%]

0.010

0.005

0.000

-0.005
0 200 400 600 800 1000
Time [s]
 B4131 Test 12 - CO2

0.10

time vs CO2% ceiling

time vs CO2% mid
0.08

0.06
CO2 [%]

0.04

0.02

0.00
0 200 400 600 800 1000
Time [s]
 B4131 Test 12 - O2

22.0

time vs O2% ceiling

time vs O2% mid
21.5
O2 [%]

21.0

20.5

20.0
0 200 400 600 800 1000
Time [s]
 B4131 Test 13 - CO

0.016

0.014

0.012

0.010

0.008
CO [%]

0.006

0.004

0.002

0.000
time vs CO% ceiling
-0.002 time vs CO% mid

-0.004
0 200 400 600 800 1000 1200
Time [s]
 B4131 Test 13 - CO2

0.10

time vs CO2% ceiling

time vs CO2% mid
0.08

0.06
CO2 [%]

0.04

0.02

0.00
0 200 400 600 800 1000 1200
Time [s]
 B4131 Test 13 - O2

22.0

time vs O2% ceiling

time vs O2% mid
21.5
O2 [%]

21.0

20.5

20.0
0 200 400 600 800 1000 1200
Time [s]
 B4131 Test 14 - CO

0.002
time vs CO% ceiling
time vs CO% mid

0.001
CO [%]

0.000

-0.001

-0.002
0 50 100 150 200 250 300
Time [s]
 B4131 Test 14 - CO2

0.06
time vs CO2% ceiling
time vs CO2% mid
0.05

0.04
CO2 [%]

0.03

0.02

0.01

0.00
0 50 100 150 200 250 300
Time [s]
 B4131 Test 14 - O2

22.0

time vs O2% ceiling

time vs O2% mid
21.5
O2 [%]

21.0

20.5

20.0
0 50 100 150 200 250 300
Time [s]
 Appendix C

Fax from Siemens on response of the CSC existing detector and the aspiration
smoke detection device