FlameGard 5 MSIR Modules
Download
Report
Transcript FlameGard 5 MSIR Modules
Multi-Spectral Infrared Flame Detector
with Neural Network Technology
EVERY LIFE HAS A PURPOSE…
Introduction
Flame Detection Applications
Existing Technologies
Flame Detection Requirements
Main Competitors
Key Customer Concerns
Neural Network Technology
FlameGard 5 MSIR Flame
Detector
Applications
Some of the most challenging
industrial environments include:
Oil / Gas Production
(Offshore / Onshore)
Oil / Gas Pipeline (Distribution /
Production / Transmission)
Chemical Plants
Electric Power
Aviation / Aerospace
Typical Industrial Flame Sources
The most common potential ignition sources:
Alcohols
Chemicals
Diesel
Gasoline
Kerosene
Jet Fuels
LNG / LPG
n-Heptane
Solvents
Textiles
Industrial Flame Detection Requirements
Flame Detector Performance Criteria
Detection Range
Field of View (FOV)
Response Time
Optical Integrity Check / Continuous Optical Path Monitoring (COPM)
Operating Temperature
Communication Capabilities
False Alarm Immunity
Detector must be immune to a multitude of IR signals, including:
Random motion
Modulation of heated surfaces
Hot air flow, arc welding
Reflection off water surfaces
Other related environmental nuisances which may result in false alarms
Key Customer Concerns
False Alarm Immunity
Uninterrupted 24/7 operations without shut-down for false alarms
Eliminates the high cost of emergency response
Materials, Clean-up, Resources, Production Time
Field of View / Detector Range
Precision uniform sensing for protection and reliability with no blind spots
Wider field of view at greater distances requires fewer detectors and
reduces costs (acquisition, installation, etc.)
Responsivity
Fast detection protects people; minimizes damage to equipment / facilities
Rugged Design
Must operate in extreme environments with reduced maintenance costs
Recognizing False Alarm Issues
Optical flame detector manufacturers have attempted to
resolve false alarm issues by using:
Multiple Sensors
UV, IR, Visual
Advanced signal processing techniques:
Correlation
Frequency Analysis
Periodicity Check
Ratios Analysis
Threshold Crossing
Expert System vs. NNT System
Current optical flame detection systems
Utilize expert decision systems, a set of classification conditions defined
by direct human analysis of experimental data
Generating accurate rules for classifying flames from false alarms
is extremely difficult and time consuming
Neural-Network Technology (NNT) based optical flame
detection systems
Utilize artificial neural networks (ANN), a pattern recognition mechanism
derived from statistical analysis of experimental data
Development time and effort is shifted from expert to computer
Superior discrimination of flames from false alarms is provided by
computer optimization to lower classification errors
The FlameGard 5 MSIR was exposed to over 450 flame and
non-flame events as part of its training
MSIR / NNT Processing Modes
Training Mode
A set of processing modules running on a computer, separate
from the flame detector, learns patterns from input data and
establishes coefficients of correlation to target flame or
false-alarm condition
Detection Mode
A simple arithmetic computation uses input sensor data and
constant correlation coefficients trained on archived sensor data
to decide on presence or lack of flame
The quality of decision scheme is independent of its complexity
and embedded system resources
MSIR / NNT Processing System
MSIR / NNT Waveforms
Wave Forms for the
Four Wavelengths
ANN Output
MSIR / NNT Flame & False Alarm
Testing Apparatus
MSIR / NNT Training Data Collection
& Testing
n-Heptane Flame
Burning in 1 ft x 1 ft Pan
Grade 87 Gasoline
Burning in 1 ft x 1 ft Pan
FlameGard 5 MSIR Flame Detector
MSIR with NNT
FlameGard 5 MSIR Mechanical Dimensions
FlameGard 5 MSIR Connectivity
FlameGard 5 MSIR Modules
Base Housing
Module
Optical Housing
Module
FlameGard 5 MSIR Cross Section
FlameGard 5 MSIR Mounting Accessories
FlameGard 5 MSIR COPM –
Every two minutes
Green and Red LED
Indicators
FlameGard 5 MSIR Modules
Electro-Optical & Termination Modules
FlameGard 5 MSIR Modules –
Dip Switch Settings
#
Option
On/Closed
Off/Open
1
High Sensitivity
1&2
2
Medium Sensitivity
1
2
3
Low Sensitivity
2
1
4
0-second ALARM Time Delay
3&4
5
8-second ALARM Time Delay
4
6
10-second ALARM Time Delay
7
14-second ALARM Time Delay
8
ALARM Non-Latching
9
ALARM Latching
10
WARN Non-Latching
11
WARN Latching
6
12
ALARM Normally Energized
7
13
ALARM Normally De-Energized
14
WARN Normally Energized
15
WARN Normally De-Energized
3
3&4
3
4
5
5
6
7
8
8
FlameGard 5 MSIR Modules –
Wiring Terminals
Terminal Block – P2
Terminal Block – P1
Pin
Description
Pin
Description
1
CAL_ IO
10
CHGND/CHASGND
(Chassis Ground)
2
COM2-/DATA2-
9
GND/COM
3
COM2+/DATA2+
8
+24 VIn
4
RLY_10
(Relay Reset)
7
0-20mA
5
ALM 2
6
COM1-/DATA1-
6
ALM 1
5
COM1+/DATA1+
7
ALM C
4
TEST_10
(Test Mode)
8
WARN C
3
FLT C
9
WARN 1
2
FLT 1
10
WARN 2
1
FLT 2
Arc Welding Immunity
Detection
Range
Immunity
Distance
MSIR/NNT
230 ft
> 5-15 ft
IR3 –
Competitor A
210 ft
> 40 ft
IR3 / CCTV –
Competitor B
200 ft
> 40 ft
Technology
Field of View Comparison
Field Of View with Sensitivities
Detection Area Coverage
FlameGard 5 UV/IR
Detection Area Coverage
FlameGard 5 MSIR
Detection Area Coverage
FlameGard 5 MSIR
FlameGard 5 Test Lamp
35 Feet
(10.7 meters)
FlameGard 5 MSIR Features / Benefits
Features
Benefits
Advanced multi-spectral IR sensor
Widest Field of View
Neural Network Technology (NNT)
Superior false alarm immunity and range
COPM
Checks optical path integrity, detectors,
electronics
Event Logging
Records time, date and type of event
Test Lamp (Auto-Detect or Manual)
Tests instrument response and functionality
HART and Modbus Communication
Options
Provides complete status and control
capability in the Control Room
FlameGard 5 MSIR Highlights
Sets a new industry benchmark
Superior false alarm immunity: Arc weld as close as 5-15 ft
Greater detection range: 230 ft max
Expanded field of view: 100 degrees @ 100 ft
Test Lamp: Test the unit from 35 ft
MTBF: 150,000 hrs
FM Certified to IEC 61508 (SIL 3 suitable)
High quality stainless steel package
MSIR/NNT provides highest levels of flame detection
with respect to performance, safety, reliability and value
Approvals: FM, CSA, ULC, ATEX, CE Marking