Embed Size (px)
DESCRIPTION
Citation preview
Smoke Detection Solutions for Challenging Environments
Trade-offs for Optimum Protection
Challenging Environments
Available Technologies
Open-area Smoke Imaging Detection and Applications ─ Technology overview
─ Field-proven applications
Industrial Very Early Warning Smoke Detection ─ Detection approaches
─ Deploying ASD in industrial environments
─ Field-proven applications
─ ASD criteria matrix
Q & A
Agenda
2
Optimum Protection Requires a Balance
3
Fire Triangle
Detection Triangle
Challenging environments complicate achieving optimum protection
Provides the highest level of detection to ensure protection of… ─ Life
─ Property
─ Business continuity
─ Environment
Without false alarms that ─ Create disruptive false or nuisance alarms
─ De-sensitize personnel and occupants to alarms
─ Ultimately leading to disregarding alarms or disabling safety systems
─ Place an extra burden on local first responders
At an affordable cost ─ Initial installation cost
─ Long term operating, servicing, and testing costs
The Detection Triangle
4
A ‘Challenging Environment’ is a space where:
Detection is difficult due to:
─ Environmental conditions that cause false and nuisance alarms
─ Environmental conditions that shorten detector life
─ High and variable airflows
─ High ceilings
Detector sensitivity and longevity are compromised in
‘harsh’ environments due to:
─ Dirt and dust
─ Temperature extremes
─ High EMI
─ Moisture, fog, steam, water condensation
─ Birds and insects
─ Building movement
─ Sunlight
─ Detector location
─ Toxic and corrosive gases
Challenging Environments
5
A ‘Challenging Environment’ is a space where:
Maintenance is difficult
─ Extreme environmental conditions
─ Inaccessible areas
● High ceilings
● Roof or floor spaces
● Within elevator shafts
● Equipment racks
─ Restricted access
● High security areas
● Production areas
● 24/7/365 operation
Unobtrusive detection is required
─ To prevent vandalism or tampering
─ Not disruptive to architectural designs
─ Preservation of historic buildings and artifacts
Evacuation may be challenging
─ High concentration of large number of people
─ Exit paths are restricted
─ Occupants require assistance
Challenging Environments
6
Detection Technologies Overview in Challenging Environments – Large Open Spaces
7
Relative ranking: 5 = best / most desirable
Driver ASD OSID Beam VSD Laser Spot Flame L. Heat
Sensitivity to smoke 5 3 2 2 4 2 1 1
Prevent damage to structure 5 3 3 3 3 2 1 1
Prevent damage to contents (equip,
product, etc.) 5 3 3 3 3 2 1 1
Foreign particle rejection (dust, fog,
water, etc) 5 4 2 3 2 2 4 5
Foreign object rejection 5 5 2 4 5 5 5 5
Resistance to building movement 5 4 2 4 5 5 5 5
Operation in all lighting conditions 5 4 2 2 5 5 3 5
Equipment Cost 3 5 5 1 2 3 1 3
Installation costs 3 5 4 4 2 2 4 3
Maintenance costs 4 4 2 4 2 2 4 5
Total 45 40 27 30 33 30 29 34
New Advancements in Open-area Detection
8
Open-area Smoke Imaging Detection
Technical obstacles overcome
Applications
Industrial Very Early Warning Smoke Detection
Detection technology overview
New advancements in VEWSD for industrial
environments
Field-proven applications
ASD requirements matrix
False alarms due to…
Dirt
Dust
Steam
Insects
Banners
Fork lift trucks
9
Challenges with Beam Detectors
Ladders
Watermist Spray
Building movement
Building vibration
Sunlight
Reflected sunlight
The improvements over traditional beam detectors stem from four core design ideas:
Dual-wavelength light frequencies
Digital imaging vs. photodiodes
A unique method for aligning
Smarter algorithms
10
What Makes Dual-wavelength Different
Large particle obscuration rejection
Use of UV & IR in specific wavelengths assist in evaluating whether
obscuration is caused by small particles in the physical size resembling
smoke or from larger particle (i.e. dust, steam, insects, fork-lift, ladders,
etc.)
Tolerance to building shift and vibration
The multiple pixels of a CMOS imaging chip as opposed to a photodiode
along with the uniquely coded light beam from an emitter provides the
ability to track the position and tolerate movement
Foreign light intrusion
The Imager is fitted with a dyed glass filter, designed to be almost opaque
except to all but two wavelengths of interest
11
Overcoming Conventional Challenges
Emitter
Obstruction
Dust
Smoke Imager
UV beam
IR beam
Active emitter LEDs transmit wide beam IR and UV to the Imager
– IR and UV have differing wavelengths
– Respond differently to smoke
Relative strengths of the UV & IR are compared
– Detect smoke
– Discriminate against particulates that cause nuisance alarms on traditional beams
Software locates illuminated pixels on the CMOS imager
– Each emitter is uniquely coded
Imager software tracks building movement
– No controlled motor drives
12
Dual-wavelength vs. Single IR Beam
Imager versus photodiode
One imager equals 100,000’s
of photodiodes
An Imager locates and tracks
the position of an emitter
anywhere in its field of view at
pixel level
Can operate reliably in all
lighting conditions; bright day /
sunlight to total darkness
Emitters
Actual view from a 45 degree imager with 4
emitters in a 13,500 sq ft area
13
Powerful Benefits
Simple and easy installation using a unique Laser Alignment Tool
─ Requires only rough alignment due to the wide angle of view of the
imager(s)
─ Up to 70% time saving compared to traditional beams
Beam length up to 492 ft
─ Outperforming traditional beams by up to 50%
14
Installation & Commissioning
15
OR
Installation & Commissioning
Real Field Challenges
Overcome by Open-area
Smoke Imaging Detection
Long Distances
Reliably detecting at 497 ft
Allows for 20% setting at 430 ft
Aligned through metal support
structure of the roof
17
Industrial Site
18
6”/15 cm
End-user was considering
linear heat cable for
‘controlled’ burn down…
Metro Station and Tunnel
5,85 “/15 cm
75ft/23 m
69ft/21m
187 ft/57 m
High ceiling
Low ceiling
46 ft/14 m
See effect of whirling dust
when metro comes into
the station
19
Cathedral York Minster
5,85 “/15 cm
Reliably detecting on 72
feet high on distance of
400 ft
Fire tests
20
5,85 “/15 cm
Effect of six Viking-style torches.
Nuisance or alarm, now you
know and choose
21
Cathedral York Minster
Semi-open Warehouse
22
Example of the effect of condensation
on the emitter
No differentiation between IR and UV
and hence no alarm nor faults
Chicken Farm
5,85 “/15 cm
Spikes with up to 20% when
chickens run around.
Estimated time between
cleaning of lenses 2-3 month
23
-10
0
10
20
30
40
50
60
16/0
2/2
012 0
0:0
0
16/0
2/2
012 0
0:0
0
16/0
2/2
012 0
0:0
0
16/0
2/2
012 0
0:0
1
16/0
2/2
012 0
0:0
2
16/0
2/2
012 0
0:0
3
16/0
2/2
012 0
0:0
4
16/0
2/2
012 0
0:0
4
16/0
2/2
012 0
0:0
4
16/0
2/2
012 0
0:0
6
16/0
2/2
012 0
0:0
6
16/0
2/2
012 0
0:0
7
16/0
2/2
012 0
0:0
7
16/0
2/2
012 0
0:0
8
16/0
2/2
012 0
0:0
9
16/0
2/2
012 0
0:0
9
16/0
2/2
012 0
0:1
0
16/0
2/2
012 0
0:1
0
16/0
2/2
012 0
0:1
2
16/0
2/2
012 0
0:1
2
16/0
2/2
012 0
0:1
2
16/0
2/2
012 0
0:1
4
16/0
2/2
012 0
0:1
4
16/0
2/2
012 0
0:1
4
16/0
2/2
012 0
0:1
6
16/0
2/2
012 0
0:1
6
16/0
2/2
012 0
0:1
6
16/0
2/2
012 0
0:1
8
16/0
2/2
012 0
0:1
8
16/0
2/2
012 0
0:2
0
16/0
2/2
012 0
0:2
0
16/0
2/2
012 0
0:2
0
16/0
2/2
012 0
0:2
0
16/0
2/2
012 0
0:2
2
16/0
2/2
012 0
0:2
2
16/0
2/2
012 0
0:2
4
16/0
2/2
012 0
0:2
4
16/0
2/2
012 0
0:2
4
16/0
2/2
012 0
0:2
6
16/0
2/2
012 0
0:2
6
16/0
2/2
012 0
0:2
6
16/0
2/2
012 0
0:2
6
16/0
2/2
012 0
0:2
8
16/0
2/2
012 0
0:2
8
16/0
2/2
012 0
0:2
8
16/0
2/2
012 0
0:2
9
16/0
2/2
012 0
0:3
0
16/0
2/2
012 0
0:3
0
16/0
2/2
012 0
0:3
0
Series1
Series2 IR Att
Carton Printing Facility
24
Water mist is sprayed every few
minutes to maintain a humid
environment
This happens 24/7, the water mist is
causing false alarms on installed
beams
UV Att
Recycling Plant
5,85 “/15 cm
Estimated time between
cleaning of lenses 1 month
25
Recycling plant 2
5,85 “/15 cm
Providing the best solutions means
also knowing what does NOT work
and exactly WHY
-20
0
20
40
60
80
100
120
23/12 10:48 23/12 12:00 23/12 13:12 23/12 14:24 23/12 15:36 23/12 16:48 23/12 18:00 23/12 19:12 23/12 20:24 23/12 21:36 23/12 22:48 24/12 0:00 24/12 1:12 24/12 2:24 24/12 3:36 24/12 4:48
Att
en
uati
on
(%
)
Date/Time
IR Atten
UV Atten
26
Direct Sunshine
5,85 “/15 cm
Disregarding the ‘avoid East – West
direction’ recommendations, Dual
Wavelength only generates a
saturation fault when exposed directly
to the sun.
27
New Advances in Open-area Detection
28
Open-area Smoke Imaging Detection
Technical obstacles overcome
Applications
Industrial Very Early Warning Smoke Detection
Detection approaches
Deploying ASD in industrial environments
Field proven installations
ASD requirements matrix
Technology Approaches to Very Early Warning Smoke Detection
Detection Approaches
− Absolute
− Adaptive
Impact on Detector Performance
− Fixed thresholds
− Drift compensation / dynamic thresholds
− Dust rejection
System Integrity
29
Continued…
Xtralis detectors use clever patented methods to ensure all optical
surfaces are free from contamination throughout their life
Clean air barrier protects optical surfaces
Eliminates detector drift
Consistent & predictable alarm thresholds
Field adjustable, fixed detection thresholds
Absolute Detection
1. Establish background levels
2. Select alarm thresholds
Dirty Environments Establishing Fixed Alarm Thresholds
Fire 1
Alert
Action
Background
Fire 2
Smoke test
31
Mathematic algorithms used to compensate for detector drift caused by
non smoke particles entering the smoke chamber or electrical
interference induced into the detection system
Short term events
Long term drift
Drift compensated alarm thresholds
Adaptive Detection - Drift Compensation
32
Drift Compensation Risks
Alert Level
95 min
145 min
Drift compensation can result in
significant detection delay
1
2
3
Detector output
33
Approaches used to compensation for dirty backgrounds
─ System pipe network design
─ Background compensation
─ Filtration
─ Maintenance
─ Particle size rejection
Particle size
Smoke Dust
Part
icle
Num
ber
Smoke detector patents for particle size discrimination
US 7564365: Marman & Eggers - GE Security
US 7483139: Powell - Kidde IP Holdings
34
Dust – an Environmental Challenge
The Conundrum:
Rejecting smoke, dust, or both?
Particle size
Smoke Dust
Part
icle
Num
ber
Typically
Smoke 0.1 – 6 m
Dust 1 – 100+ m
Mulholland, G.W. , SFPE Handbook of Fire Protection
Engineering 2nd Ed, Ch 15, Section 2.
Lide, D.R. (1994) , Characteristics of Particles and
Particle Dispersoids, Handbook of Chemistry and
Physics, 75th Ed.
35
Dust Rejection – an Environmental Challenge
Dust rejection may result in reduction in
detector sensitivity
1 ”A comparison of Aspirated Smoke Detectors. Honeywell
FAAST Detectors and Xtralis VESDA VLF LaserFocus”
Conducted by Packer Engineering, Inc and The Fire
Testing Evaluation Center at The University of Maryland,
College Park, Nov 2010
Adaptive vs Absolute Detection
36
Adaptive Detection
Absolute Detection
Dust Rejection – the Consequences
Sample Holes
Return-air Grill
1 NA: No Alarm within 300sec
Test ID
Response Time (sec)
Absolute
Detection
Adaptive
Detection
1 188 sec – Alert NA1
2 220 sec – Alert NA
3 132 sec – Alert NA
4 178 sec – Alert NA
Test
ID
Test Fire
Time to Alarm (sec)
Absolute Detection Adaptive Detection
Alert Action Fire Alert Action 1 Action 2
5 Timber -smoldering 590 787 NA1 799 NA NA
6 PU foam -smoldering 628 876 NA NA NA NA
7 PU foam -flaming 140 167 198 190 NA NA
8 Heptane - flaming 78 143 266 175 NA NA
9 Paper – smoldering 660 NA NA NA NA NA Tests witnessed by Centre for Environmental Safety
and Risk Engineering (CESARE), Victoria University,
Melbourne, AU 37
Adaptive vs. Absolute Detection
Electrical integrity
Detection chamber
Pipe network flow monitoring
- Ensure that it cannot be by-
passed
Filter life monitoring (currently on
the market)
- No monitoring
- Time based
- Particle counting
- Through filter flow monitoring
The FACP says these detectors are
fully functional
38
System Integrity – Critical Functions Must be Monitored
System Integrity – time based filter monitoring
Unsupervised Detection Failure 39
Sensitivity vs. Filter Loading Cycle
Intelligent Filtration
Larger x% of flow passes through
HEPA filter
Smaller y% of flow passes
unfiltered
Flows recombined
to enter detector at lower
obscuration than original
Pipe Flow Splits into x% and y%
Flow Sensors
Air coming directly from sampling pipe
(total airflow)
Reduce contaminates
entering detection chamber
Innovative flow partitioning
and HEPA filtration
Continuous monitoring of
filter and airflow
Optimum detector sensitivity
over detector life.
Very Early Warning Smoke Detection in Industrial Environments
Perceptions of ASD
ASD Comfort Zone (Roots in Data Center and Telecommunications)
Uncomfortable Zone (Industrial)
Dirty Clean
Designed for the cleaner environments
Industrial applications beyond ASD capabilities
Market skepticism still exist
Historically ASD system designs in all industrial
environments required system application engineering 42
Improved general protection
− Hardened product enclosure – IP/NEMA rated
− Eliminate need for secondary enclosures
Longer pipe run to accommodate larger facilities
Greater visibility of status LED
Improved filtration
− External filtration can stress system maintenance
− Eliminate need for external filters
Field serviceable modules
− Less down time
− Lower spare part inventory
Lower long term operating costs
Establish standard to quantify contamination resistance
43
Industrial Applications Require More from ASD
IP54 Enclosure
Intelligent filtration
1,200 ft. pipe runs
High intensity status LED
Field serviceable – modular
construction
44
Fit for purpose Industrial ASD
Enhanced filtration & IP-rated enclosure
can reduce Total Cost of Ownership by 40 to 60%
In-line Filters
Detectors inside protective enclosures No protective enclosures required
No In-line Filters
45
New Approach to ASD in Industrial Environments
Filters
Aspirator
Detection chamber
Fewer spare parts – less down time
Electronics
46
Modularity Delivers Lower Total Cost of Ownership
Power Plant – USA
Steel Mill – USA
47
Industrial Installations
Illovo Sugar Mill – South Africa
Cook Colliery – Australia
Boiler Feed Pump
ASD Protected from extreme
heat via pipe network
Battery rooms
ASD + hydrogen gas detection
48
Power Plant
300 foot tunnel under electric reheat
furnace
Furnace utilizes nitrogen protective
atmosphere – risk of oxygen
depletion (asphyxiation)
Cable trays – risk of fire
Solution: Industrial ASD + Oxygen
49
Steel Mill – Tunnel Protection
ASD Detector in
Protective Enclosure
Industrial ASD
Detector
50
Illovo Sugar Mill – Pipe Network Design & Installation
51
Cook Colliery – Underground Mine
52
Cook Colliery – Pipe Network Design
53
Cook Colliery Installation
Clean
Dirty
High
Contamination
Computer / Control / Clean
Rooms / Data / Telco
Switch Rooms /
Warehousing
Conveyor / Mining
/ Other
Switch / Substation / Warehousing / Factory / Recycling / Processing /
Manufacturing / Power Gen / Conveyor / Mining , High Dust, Etc.
Additional considerations
may be required
In-line filters (Optional)
Contamination
Bagasse / Coal/
Underground
- Various Manufacturing and
Warehousing Applications -
Fertilizer / Paper / Steel /
Tunnels / Etc.
Semi Conductor / Communication /
Control / Data
May require
some
applications
engineering
In-line filter
Evolution of ASD Applications
55
ASD requirement matrix
De
tec
tor
1
De
tec
tor
2
De
tec
tor
3
- Consistent Alarm Thresholds over detector life
- Determinate alarm thresholds (% obs/ft or %obs/m)
- Ability to compensate for background levels
- Flow system monitoring that cannot be by-passed
- Active filter life monitoring
- Quantified contamination resistance (for industrial)
Proposed contamination test
Dust type: per ASHRAE 52.2
Dust loading density: 16mg per cubic meter
Test cycle: (dust - clean air - smoke)
Dust duration: 3.5 hours
Clean air: 30 minutes
Smoke test: smouldering cotton - two exposures
Number of test cycles: 10
Allowable sensitivity drift: <15%
Proposed ASD Criteria Selection Matrix
Q & A
56
