The challenge of combustible gas and fire safety continues every day at thousands of industrial process facilities throughout the world, but it is worse in many developing countries. Hundreds of preventable accidents are reported every year – many of them with the tragic unnecessary loss of employee lives. There is hope, however, because the latest generation of gas and flame safety detection systems is smarter than ever and can provide a layered human sensory approach that is more comprehensive to protect people and facilities.
Plant Protection Challenges
The primary sensing technologies utilized in the detection of combustible gas leaks and flames at industrial process plants are: catalytic bead, point infrared (IR), open path IR, ultrasonic and optical. These technologies are all well-known with more than a decade of proven performance in the field. All of them have their unique advantages depending on the application environment. All of them are individually susceptible to false alarms under certain conditions. No single method of gas and flame detection is foolproof.
Every industrial process plant has a variety of potential gas leaks and flame leak sources, including leaking tanks, pipes, valves, pumps, etc. The avoidance of false alarms is also important because they result in unnecessary process or plant shutdowns, slowing production and requiring time-consuming reviews, paperwork, and reporting. False alarms can over time provide employees with a false sense of security because they become complacent if alarms go off for no apparent reason and then just choose to ignore them.
Detecting dangerous gas leaks and flames reliably is a challenge using any single one of the conventional technologies. For example, IR detectors can’t detect hydrogen gas because hydrogen doesn’t absorb IR energy. In another example, while a pressurized pipe gas leak can create an ultrasonic noise, so can other pieces of equipment that can trigger false alarms in ultrasonic detectors. Reflections or heat rising off tanks and other shiny surfaces on hot days can fool optical flame detectors.
The Human Sensory Model
With all the challenges that the best sensing technologies face, it’s not surprising that a new strategy in gas and flame protection is emerging for industrial plants. What if you combined all of the gas and flame detection technologies together and then layered them where they fit best in terms of their reliability in each unique plant layout?
If you examine today’s gas and flame detection sensing technologies, they mimic the senses of the people who invented them. Catalytic bead detectors “sniff” gases for example, infrared and optical type sensors “see” gases and flames, and ultrasonic sensors hear “gases”. What if some of these detectors behaved more like people, reacting based on their intelligence and retained past memories?
A new strategy of layering detector technologies throughout the plant where they fit best in terms of their reliability achieves a human sensory chain of plant defense against hazardous gases and flames. To learn more about this new model of human sensory gas and flame detection, let’s look at each type of sensing technology, discuss how they work and then look at how artificial intelligence is being applied today to advanced gas and flame detection.
Catalytic Bead (CB). The operating principle of catalytic gas detectors employs catalytic combustion to measure combustible gases in air at fine concentrations. As combustible gas oxidizes in the presence of a catalyst, it produces heat and the sensor converts the temperature rise to a change in electrical resistance, which is linearly proportional to gas concentration. A standard Wheatstone bridge circuit transforms the raw temperature change into a sensor signal.
Point Infrared (PIR). Infrared gas detectors use two wavelengths, one at the gas absorbing “active” wavelength and the other at a “reference” wavelength not absorbed by the gas; neither wavelength is absorbed by other common atmospheric components such as water vapor, nitrogen, oxygen, or carbon dioxide. In point IR detectors, the concentration of hydrocarbon gas is measured via the infrared absorption of an optical beam known as the active beam. A second optical beam, known as the reference, follows the same optical path as the active but contains radiation at a wavelength not absorbed by the gas.
Open Path Infrared (OPIR). The OPIR detection path of the IR beam is expanded from less than 10 centimeters, typical of point IR detectors, to greater than 100 meters. These devices can use a retro-reflector or separate IR transmitters and receivers housed in different enclosures. There are OPIR detectors available that monitor in both the LEL-m and ppm-m ranges to detect both small and large leaks. They cover large open areas, along a line of several potential leak sources such as a row of valves or pumps and also for perimeter monitoring of leaks.
Ultrasonic (UGLD). In comparison to conventional gas detectors that measure % LEL, advanced ultrasonic gas leak detectors with neural network technology (NNT) include pattern recognition capability that responds to the ultrasonic noise created by a pressurized gas leak. This ultrasonic noise provides a measurement of the leak rate and establishes warning and alarm thresholds. Gas does not need to reach the sensing element as the detector “hears” the gas leak. They are best suited for outdoor installations and indoor spaces with high ventilation rates.
Ultraviolet/Infrared (UV/IR). By integrating aUV optical sensor with an infrared (IR) sensor, a dual band flame detector is created that is sensitive to the UV and IR radiation emitted by a flame. The resulting UV/IR flame detector offers increased immunity over a UV-only detector, operates at moderate speeds of response, and is suited for both indoor and outdoor use.
Multispectral Infrared (MSIR). Advanced multispectral infrared (MSIR) flame detectors combine multiple UV/IR sensing arrays with neural network intelligence. They provide pattern recognition capabilities that are based on training to differentiate between real threats and normal events thus reducing false alarms. MSIR technology allows area coverage up to six times greater than that of more conventional UV/IR flame detectors.
Artificial Neural Networks (ANN)
The concept of artificial neural networks (ANN) is now being applied by safety monitoring system manufacturers to develop their own neural network technology (NNT) to improve conventional gas and flame detectors.
The concept of NNT is based on the human brain, and this technology is now being applied to gas and flame detection. Detectors equipped with NNT intelligence provide a more reliable solution because they can eliminate many false alarm sources while improving overall detection.
During the 1940s, computer researchers developed the first conceptual model of an artificial neural network after studying the human brain to solve certain kinds of problems that are easy for humans, but difficult for computers – otherwise known as pattern recognition. There are today a variety of applications of neural networks, some of which are at work in gas and flame detectors. They include:
• Pattern recognition
• Signal processing that can filter out irrelevant data
• Controls that manage decisions
• Soft sensors that analyze a collection of many measurements
• Anomaly detection – the ability to generate output when something occurs that doesn’t fit patterns thus issuing alerts when something is amiss
NNT, which is based on ANN, is in essence an artificial intelligence. A key advantage of this technology is its ability to learn. It learns through a type of apperceptive process; meaning the comprehension or assimilation of something such as a new idea can then be related in terms of previous experiences or perceptions. NNT operates similarly and is much like a human mind in the way that it enables a person to recognize a face from the distant past. For example, the brain facilitates recognition by matching a face with an image stored as a memory.
Next generation gas and flame detectors, just like the human brain, each have thousands of pieces of data stored in their memories from hundreds of gas leak, non-gas leak, flame and non-flame events that have been observed in the past.
These detectors have been trained through NNT intelligence to recognize an actual gas leak or flame based upon that data, and they make decisions about whether they are detecting an actual gas leak or flame, even if they have not seen that exact pattern in the past.
Conclusions
The human sensory model of multi-layer gas and flame protection builds a chain of defense to protect industrial plants. This model allows process and plant engineers to add layers of protection where it makes sense to increase overall system reliability.
By Ardem Antabian, OGP Segment Manager
MSA – The Safety Company
www.msagasdetection.com
Print this page
