International Road Tunnel Fire Detection Research Project - istss

International Road Tunnel Fire Detection
Research Project
Z. G. Liu, A. Kashef, G. Crampton and G. Lougheed,
National Research Council Canada, Canada
Daniel T. Gottuk, Hughes Associates, Inc.
Kathleen H. Almand, Fire Protection Research Foundation

Project Background
The tunnel fire detection project was initiated in
1999 at the request of Port Authority of New
York and New Jersey and Boston Fire
Department.
• Phase I – literature review;
• Phase II – full-scale experiments.
Phase I – literature review was completed in
2003.
Phase II – full scale testing was initiated in 2006
by FPRF and NRC.

Project Objectives
Investigate performance of current fire
detection technologies;
• Detection capability for realistic fire scenarios
• Reliability in tunnel environments
Provide technical data for developing
standards and guidelines;
Assist in optimization of technical
specifications and installation requirements
for tunnel fire protection.

Tasks & Methodology
Task 1: Identify technologies and develop test
protocols
Task 2: Fire tests in a laboratory tunnel
Task 3: Computer modeling
Task 4: Fire tests in an operating tunnel in Montreal
Task 5: Environmental tests in Lincoln Tunnel in New
York
Task 6: Demonstration fire tests in Lincoln tunnel in
New York
Task 7: Ventilated fire tests in a laboratory tunnel

Task 1: Detection Technologies,
Fire Test Protocols

Detectors in Project
Technology
Linear heat
System
No.
D-1L1
System Information
Fiber optic heat detection system
Flame D-3F1 Multi-IR flame detector
CCTV
Spot heat
Smoke
D-2L2
D-4C1
D-5C2
D-6C3
D-7H1
D-8H2
D-9S1
Analogue (co-axial cable) heat
detection system
Visual based flame/smoke detector
Visual based flame/smoke detector
Visual based flame detector
Frangible bulb heat detector
Rate-anticipation heat detector
Air sampling system

Performance Criteria
for Detection Systems
Detection capability
• Detecting fire scenarios;
• Locating fire position;
• Monitoring fire development.
Detection reliability
• Reliable in detecting a fire – low false alarm rate;
• Reliable in working in harsh environment – limited
maintenance requirements.

Tunnel Fire Scenario #1
Pool fires: fast growing fires (gasoline, propane)
• Open space fires (125 kW);
• Fires underneath a vehicle (125 ~ 3400 kW);
• Fires behind a large vehicle (650 ~ 3400 kW)

Tunnel Fire Scenario #2
Stationary vehicle fires: slowly growing fires (wood,
foam, gasoline)
• Engine compartment fires (~2000 kW)
• Passenger compartment fires (100~1500 kW)

Tunnel Fire Scenario #3
Moving vehicle fires: fast moving fires (~150 kW)
• Fires moving in two directions: facing and away from
detectors
• Two speeds: 25 km/h, 50 km/h

Task 2&7:
Fire Tests in a Laboratory Tunnel

Detectors in Tests
Schematic of detection systems in the tunnel
CCTV 1
Linear
cables 1&2
spot 1
spot 2
smoke
CCTV 2
Flame
CCTV 3
2.5 m
5 m
4.1 m
Wind
Fire detectors faced openings of the tunnel;
The wind was toward detectors;
Three air velocities: 0 m/s, 1.5 m/s, 3 m/s

Task 4: Field Fire Tests in
Montreal Tunnel

Detectors
– D-1L1: fiber optic linear heat detection system
– D-2L2: co-axial cable linear heat detection systems
– D-3F1: IR3 optical flame detector
– D-4C1: visual-based CCTV flame smoke detector
– D-5C2: visual-based CCTV flame/smoke detector
– D-6C3: visual-based CCTV flame detector

Fire Detectors & Scenarios
CCTV 1
CCTV 2
Flame
Linear
cables
1&2
Jet fan
CCTV 3
55 m
FP1
FP3
30 m
FP2
30 m
FP4
Wind
4-lanes
Dimensions of tunnel: 600 m long,
16.8 m wide (4 lanes), 5 m high;
Four jet fans in tunnel;
Six detectors/detection systems;
Fire at four locations

Fire Tests in Montreal
Tunnel (Task 4)
Fire scenarios: small open fires, pool fires under a vehicle and
pool fires behind a large vehicle
Four wind speeds: 0 m/s, 1.3 m/s, 2 m/s, 2.4 m/s

Task 3 – Numerical
Modeling
Preliminary
Comparison
Tunnel Ventilation
Systems
Tunnel Length
Conduct CFD simulations to assist in the preparation of fullscale
experiments conducted in the laboratory tunnel facility
Conduct CFD simulations to compare numerical predictions
against the full-scale experimental data (Tasks 2, 4, 7)
Conduct CFD simulations to investigate the impact of tunnel
ventilation conditions (longitudinal, semi-transverse and fullytransverse
ventilation systems) on the development and
distribution of the temperature and smoke
Conduct CFD simulations to study the impact of tunnel length
on the development and distribution of temperature and
smoke in the tunnel

Task 3 – Comparisons
against Task 7
58
33
28
Temperature ( 'C )……
200
180
160
140
120
100
80
60
40
Temprature at Thermocouple #58
1.5 m/s
2.0-1.5m/s Slot 58
Tun2BV25 T38
Temperature ( 'C )……
300
250
200
150
100
50
0
Temperature at Thermocouple # 28
Plume fluctuations
Airflow conditions at entrance
Radiation
0 100 200 300 400
Time ( s )
2.0-1.5m/s Slot 28
Tun2BV25 T8
20
0
0 100 200 300 400
300
Temperature at Thermocouple #33
Time ( s )
250
1.0 m x 2.0 m Gasoline 3300~3400 kW 1.5 m/s 21 o C
Fire Behind
Vehicle
Temperature ( 'C )……
200
150
100
50
2.0-1.5m/s Slot 33
Tun2BV25 T13
0
0 100 200 300 400
Time ( s )

Task 5: Environmental Field Tests in
Lincoln Tunnel at New York City
Daniel T. Gottuk, Ph.D., P.E.

Program Overview
• Impacts of real tunnel environments on reliability of
fire detection systems, including their false alarm
rates and maintenance requirements investigated.
• Evaluation of 3 fire detection technologies
– Video Image Detection (VID) for flame and smoke
– Optical Flame Detection (OFD)
– Smoke Aspiration Detection
• 4 Detection systems installed and being monitored
• Long-term ( 1 year) monitoring of fire detection
systems and fire demonstration tests

Detection Systems
ID Technology System
Information
D-3F1 OFD Flame
Hardware
Location
Roadway
D-4C1 VID Smoke and Flame Tunnel Cameras
with Unit in
Admin. Building
D-6C3 VID Flame
Roadway
D-9S1 ASD Smoke
Exhaust plenum

Environment in Tunnel – 10 month period
•Outside temperatures ranged from -12°C C to 33°C.
•The snow fall was 2.5 cm and the rain fall was
19.2 cm.
•High soot and dirt levels from traffic
•Overspray from periodic washing of the tunnel
walls and ceiling with a water and soap

Task 6: Demo Fire Tests in Lincoln
Tunnel at New York City

Fire Demonstrations
• November 11, 2007
• 5 fire events
– Diesel pan fires in back of stripped-down van
– ~1 MW to 2 MW
– Burn time ~5 minutes
– Rear of vehicle toward detectors
– Flame visible through window openings (area 0.44
m 2 )
• 2 fires near NJ portal
• 3 fires near center of tunnel

Findings

Findings – Fire Effects on System
Performance

Findings – Fire Effects on
System Performance

Findings – Ventilation Effects on
System Performance

Findings - CFD
Temperature ( 'C )……
300
250
200
150
100
50
Temperature at Thermocouple # 28
Plume fluctuations
Airflow conditions at entrance
Radiation
2.0-1.5m/s Slot 28
Tun2BV25 T8
0
0 100 200 300 400
Time ( s )

Lincoln Tunnel
Environment Performance
ID Technology System Information Nuisance
Alarms
Trouble and Maintenance
Issues
D-3F1 OFD Flame 3 Numerous optical faults
D-4C1 VID Smoke and Flame 164 Multiple trouble
D-6C3 VID Flame 0
D-9S1 ASD Smoke 1

Lincoln Tunnel
Demonstration Test
Results
Test ID
Fire Location
Dis. From
Dets. (m)
Results
Demo 1 Near NJ Portal 61 No detection
Demo 2 Near NJ Portal 30 Only OFD alarmed
Demo 3 Near Center 61 No detection
Demo 4 Near Center 30 Only ASD alarmed
Demo 5 Near Cener 15 OFD and ASD alarmed

General Findings –
Lincoln Tunnel
• Dirt and grime a major factor in all types of system
performance
• Sunshine and headlights a factor in VID performance
• Direction of traffic flow (induced ventilation) a factor in
performance

Recommendations for
Further Research
• The impact of transverse or semi-transverse
ventilation, on the performance of fire
detection systems
• More studies on the development of flame and
smoke under wind conditions for VID systems
• False alarm propensity of for linear heat
detection
• Refinement of CFD modeling tools for CCTV
and flame detection

Summary
A test protocol for evaluating various fire
detection systems for tunnel protection has
been developed;
Detection capability and reliability of detection
systems in tunnel environments were
systematically investigated with full-scale tests
in laboratory and operating tunnels, as well as
computer modeling;
The strength and weakness of current
detection systems in tunnel applications have
been identified;

Summary
Technical data provided from the project can
be used for development of standards and
guidelines; improvement of fire detection
technologies and their applications in tunnels,
as well as tunnel safety.

Acknowledgements
Authors would like to acknowledge the contributions
of following organizations to the project:
• Port Authority of New York and New Jersey; Ministry of
Transportation of Quebec; Ministry of Transportation of
Ontario; Ministry of Transportation of British Columbia; City
of Edmonton; Carleton University;
• Siemens Building Technologies; VisionsUSA; AxonX/Johnson
Control; Tyco Fire Products; Sureland Industries Fire Safety;
Micropack; United Technologies Research Corporation; Det-
Tronics; J-power System/Sumitomo Electric; Honeywell;
• A & G Consultants; PB Foundations;
• members of Technical Panel and other NRC staff