Hydrocarbon Processing Copying and distributing are prohibited without permission of the publisher
Email a friend
  • Please enter a maximum of 5 recipients. Use ; to separate more than one email address.

Consider a systematic approach for safety monitoring

11.01.2013  |  Wilson, G.,  Emerson Process Management , Austin, Texas

Keywords: [detector] [gas tetector] [flame detector] [safety monitors] [IR] [Ultrasonic] [UC detectors]

Any facility dealing with bulk hydrocarbons must monitor safety conditions at all times. A leak of flammable or toxic material can occur at any time, and explosion is an ever-present threat when combustible gases are accidentally released into working areas of a plant.

Operators have a wide choice of safety equipment to ensure coverage, backed by a library full of standards and regulations, but what would be an ideal solution? The “ideal” safety instrument would be able to immediately and accurately sense flammable and toxic gases as well as flames. It would be reliable, low power and resistant to poisoning and other environmental effects. Equally important, it would be low maintenance, economical, flexible, adaptable and self-protecting. Obviously, a single instrument such as this just does not exist, but by combining the right technologies to a particular application conditions, the “ideal” level of coverage and performance can be achieved.

One way to look at protective measures is to rank them into three categories or stages: Detect > Distinguish > Defend (Fig. 1). Knowing the technologies used in each stage, their advantages and their limitations can help facilities achieve optimum solutions.

   Fig. 1 One way to look at protective measures is to rank them
  into three categories or stages: Detect > Distinguish > Defend.

Three stages of protection

 The first line of defense (and the fastest) is a solution providing wide-area coverage that can immediately detect the initial symptoms of an issue, under any environmental condition—by sensing a byproduct of an actual leak occurring. Ultrasonic gas leak detection can provide this kind of performance; it instantaneously responds to the ultrasound generated by a high-pressure gas leak, and it is practically immune to any environmental conditions. It also covers a large, spherical radius around the instrument. Yet, while this technology provides reliable detection in conditions where traditional solutions cannot, it reveals nothing about where or what exactly is leaking.

The second line of defense is fixed-point gas detection, which provides more specific information, such as: What has been released? Where is the point of the release? Is it a toxic gas or flammable gas, and at what concentration?

A fixed-point gas detector can provide that information very quickly, if not quite as fast as the ultrasonic detector, and will provide operators with an almost exact location of the release in their facility. One limitation of a traditional fixed-point gas detector is that it requires the gas to make contact with the sensing element, or, in the case of line-of-sight detectors, requires the gas cloud to enter the detection beam. Outdoor applications can be problematic for these technologies, as wind and even light airflow can move the gas away from where it can be detected, even if the leak occurs right beside a sensor. Temperature and humidity can also negatively affect the performance of fixed point detection sensors, and, of course, they all require regular maintenance and periodic sensor replacements.

Fire or flame detection is the last line of defense to avert a catastrophe. The fire has started, but hopefully it is still small, and immediate action must be taken to prevent a potential disaster. Eliminating the risk of an uncontrollable fire developing is achieved through the use of automatic suppression systems, generally combined with a voting system to ensure that false alarm sources do not trigger these systems. There are many factors to consider when selecting your flame detection technology and configuring a system. Eliminating nuisance false alarm events is of primary concern, as well as ensuring fast, reliable performance in your particular environment.

When consulting with providers on equipment choices, it is beneficial to select an equipment vendor that offers the entire spectrum of detection solutions. Not only will this simplify procurement but it makes it possible to have experts available to consult on the optimum combination of technologies, and their correct installation, within your applications.


An ultrasonic detector is an acoustic device tuned to frequencies above the range of human hearing (Fig. 2), plus associated amplifiers, signal processors and alarm circuitry. There are two acoustic sensing types available: microphone-based and piezo-electric-based sensors. A single ultrasonic detector can cover a radius of 20 meters (m) to 40 m, depending on the model required to deal with background ultrasonic noise sources, as shown in Fig. 3. This technology is almost completely insensitive to environmental conditions such as precipitation, wind and humidity. Microphone-based sensors do require periodic maintenance and calibration, while piezo-electric-based sensors are calibrated for life at the factory. Piezo-electric sensors have the further advantage of being non-consumable (never expiring) and use a ceramic seal, making them extremely robust. Both technologies come with automatic integrity testing to ensure continuous operation. A huge advantage with ultrasonic detection is that the response is essentially instantaneous even under the most extreme conditions. For added redundancy, some ultrasonic detectors are available with up to four independent sensors (Fig. 4), as well as features like field-selectable dB alarm levels and time delays and a choice of either electronic or true broadband sensor self-testing.

   Fig. 2 An ultrasonic detector consists essentially of an acoustic
  device tuned to frequencies above the range of human hearing.

   Fig. 3 A single ultrasonic detector can cover a radius of 20 m
  to 40 m, depending on the model required to deal with
  background ultrasonic noise sources.

   Fig. 4 For added redundancy,
  some ultrasonic detectors are
  available with up to four
  independent sensors, as well as
  features like field-selectable dB
  alarm levels and time delays to
  optimize installation settings.

One limitation, as noted previously, is that an ultrasonic detector provides no information on what is leaking. In addition, it is not suitable for low-pressure leaks (below 2 bar/30 psi) and can be activated by other sources of ultrasound (like pressure relief valves) if not configured correctly. These devices would not fall into a low-power consumption category, and the units themselves are slightly larger than fixed-point detectors and not as flexible to install and operate (no sensor/transmitter separation).

Manufacturers generally recommend advanced mapping services to develop the optimum instrument coverage required at a facility by factoring in potential leak sources and the complete area of coverage when evaluating background noise to establish sensor ranges and settings, combining them with fixed-point solutions.


Gas detectors are available for both combustible and toxic gases. Combustible gas sensor types include catalytic bead and infrared (IR), both point and open path, while toxic gas sensors include electrochemical and metal oxide semiconductor (MOS).

Combustible gas detectors

A catalytic bead device (Fig. 5) consists of a pair of catalytic beads with platinum coils inside, one for sensing and the other, a reference, made the same way but not catalytically active. Current through the platinum (Pt) coils warms the beads, and any combustible gas touching the sensing bead will oxidize, raising the bead’s temperature and changing the electrical resistance of the Pt coil inside linearly with gas concentration. Simple circuitry (i.e., Wheatstone bridge) compares the resistances of the sensing and reference coils to produce an output.

   Fig. 5 A catalytic bead sensor consists of a pair of catalytic beads
  with Pt coils inside, one for sensing and the other, a reference,
  made the same way but not catalytically active. Current through the
  Pt coils warms the beads, and any combustible gas touching the 
  sensing bead will oxidize, raising the bead’s emperature and 
  changing the electrical resistance of the Pt coil inside linearly
  with the gas concentration.


Catalytic bead sensors have the advantage in that they detect both hydrocarbons and non-hydrocarbon combustible gases simultaneously, which is useful where more than one gas or non-hydrocarbon gases may be present. A representative sensor can detect methane, propane, n-butane, isobutylene, hydrogen, ethane, pentane, hexane, heptane, ethylene, propylene, methanol, ethanol and more.

This technology has been proven in the field for more than a half century, and the sensors are simple, rugged and inexpensive to replace. They are resistant to difficult environmental factors like condensation and humidity, and, because they have no optics, are well suited to dusty and dirty environments. They are also available with custom calibration factors to allow some flexibility with calibration.

Considerations include susceptibility to some poisoning agents and the need for an oxygen atmosphere. Exposure to high concentrations of gas can significantly degrade the sensor’s performance; if the exposure continues long enough, it may cause the bead to overheat and fail prematurely. Some sensors are self-protecting, with a feature that automatically reduces current to the bead under these conditions, dramatically increasing the sensor’s lifespan. And compared to IR sensors, catalytic bead sensors do require more frequent calibration and periodic replacement.

IR gas detectors. These items sense gases through the principle that hydrocarbon gases attenuate certain wavelengths of IR light. The particular wavelengths are specific characteristics of the gases. In analytic work, technology referred to as nondispersive infrared (NDIR) is used to identify them. An emitter (usually some form of incandescent lamp) produces a beam of wideband IR light that is sent through the air to a wavelength-sensitive sensor. In most designs, the light from the source is split in two beams, with one going through the air to be monitored and the other through clean air as a reference.

IR sensors are available in both point and open-path versions; the point type uses an IR service life entirely within the sensor, while the open-path type sends a beam of IR light through an area to a receiver or to a reflector that sends the beam back. They are quite accurate (some using dual-light sources); are immune to poisoning and other environmental effects. Some units have automatic internal compensation for changes in temperature, humidity and light source aging, and others have long calibration intervals. Unaffected by high concentrations of hydrocarbons, most have electronic failsafe monitoring, and can function in the absence of oxygen or in oxygen-enriched atmospheres. They are economical when considering the service life of the instrument, with sensor lifetimes of eight to ten years, and typically only require a simple calibration once a year.

IR sensors are not available for all gases (hydrogen, vinyl chloride and acetonitrile, for example, cannot be detected by IR), and are an expensive technology compared to the catalytic bead. They provide a variable, nonlinear response and are configured to detect a single target gas, although some units are available with field selectable combustible gas curves to adapt to varying plant conditions.

Toxic gas sensors. The most common technologies used for sensing the presence of toxic gases are electrochemical and MOS. Electrochemical sensors are essentially tiny fuel cells. The gas to be detected enters the sensor through a gas-permeable membrane or a capillary and comes in contact with an electrolyte with a sensor electrode and a counter electrode. A reaction (oxidation or reduction) takes place between the sensor electrode and the gas, and a current is generated that is linear with gas concentration. Many sensors add a third (or even a fourth) electrode to eliminate problems with electrode polarization and to increase lifetime.

Electrochemical sensors can detect all sorts of gases, including H2S, CO, NO, NO2, NH3, SO2, Cl2, H2, HCN and HCl. Electrochemical sensors are a proven technology, known to be repeatable, reliable and accurate. Advantages include specificity of response (only the target gas is detected, although there can be exceptions), the ability to detect gases in the parts-per-million (ppm) range, and very low-power consumption.

Limitations include a tighter temperature range, limited shelf life and possible cross-sensitivity to other gases, and that the sensors are not self-protecting. One other factor to consider is sensor life. Sensors gradually deteriorate and must be replaced. One year to three years of lifetime is typical, depending heavily on environmental contaminants, temperature and humidity. Sensor life tends to be shorter under extremely hot and dry conditions.

MOS sensors work by detecting a change in electrical resistance when a target gas like hydrogen sulfide (H2S) or ammonia (NH3) is adsorbed onto the surface of a layer of metal oxide. Since response characteristics change with temperature, a built-in heater maintains a constant sensor temperature, allowing the sensor to operate over wide ranges of temperature and humidity. Advantages of the latest nano-enhanced MOS sensors include the ability to detect small concentrations of gases (less than 20 ppm), low cost, long life and fast response.

MOS sensor limitations include the power consumption needed to keep the sensing element hot (about 1 W, in some cases) and the tendency to be fooled by the presence of non-target gases like ozone, Cl2 and several others. Sensors monitoring for particularly dangerous gases like H2S use special methods to reduce sensitivity to interfering gases. In addition, some traditional MOS sensors have what most applications would consider a dangerously slow response and have a tendency to “fall asleep,” losing sensitivity unless exposed to a gas mixture on a regular basis, also known as a “bump test.” Like electrochemical sensors they also require periodic calibration (every three months). MOS sensors have a 12-minute to 15-minute warmup period and need to be powered up on location at least 48 hours before the first calibration. Finally, MOS sensors typically have a moderately higher cost.

An advancement in MOS technology, in which the surface of the oxide has nano-scale features, provides improved performance, including much faster response, no loss of sensitivity, quick recovery from high-concentration gas exposure events (a common issue with traditional MOS sensor materials), and the ability to perform well under hot/dry conditions. Some also require much less power than conventional MOS sensors.

Transmitters for use with gas detectors are available in “blind,” which generally means no display but with an LED indication, and full-character display configurations. There are transmitters available in the market with only single-channel, accepting a single sensor’s capabilities, as well as multi-channel variations that can communicate to connected sensors independently. Ideally, the selected transmitter platform would be truly universal and allow all sensor technologies to be mixed and matched as desired. Other features to consider when selecting a transmitter are low power consumption and a wide voltage range, a bright display that can be read under extreme conditions, a full-character interface with non-intrusive controls and intuitive commands, plus the availability of multiple output protocols including wireless.



Flame detectors are sensitive to energy emissions from actual flames, and are designed to minimize their sensitivity to potential sources of interference such as sunlight, electric lighting, arc welding and hot objects. Optical flame detectors utilize ultraviolet (UV) and IR technologies in various combinations.

IR detectors work at the spectrum below (longer wavelength than) visible light, and are available in near-infrared (NIR), narrowband and wideband types. NIR detectors are sensitive to wavelengths from 0.7 µm to 1.1 µm and are relatively immune to attenuating sources like water or vapor. Narrowband IR detectors look for the 4.3 µm emission characteristic of hot CO2 given off by hydrocarbon fires. Wideband IR detectors are sensitive to all wavelengths of IR, and can detect some fires that narrowband units cannot. Note: That water, ice, snow and steam all absorb IR energy and can affect all but NIR detectors.

Triple IR detectors (Fig. 6) look for three specific wavelengths: the 4.3 µm of CO2 plus two other wavelengths above and below that. They respond only if the 4.3 µm emission is sufficiently greater than the other two, which helps greatly to reduce false alarms caused by sunlight, welding, lightning, x-rays, arcs and sparks and by hot objects in the area that are not on fire. They are also resistant to the effects of known IR absorbers like rain and fog. Electronics that monitor for the 1-Hz to 20-Hz modulation (“flicker frequency”) of a flame can greatly reduce false alarms caused by heat radiation from equipment, but the detectors still may be sensitive to very strong, nearby, modulated IR energy sources. In these cases, the detector will need to be repositioned or can have shades installed to block an identified false alarm source.

   Fig. 6 Triple IR detectors look for three specific wavelengths: 
  the 4.3 µm of CO2 plus two other wavelengths above and below
  that; they respond only if the 4.3 µm emission is sufficiently
  greater than the other two, which helps greatly to reduce false alarms.

Other advantages of IR detectors include a wide temperature range, very limited calibration required, and low power requirements. Low power requirements provide significant cost savings during installation and the life of the instrument, and deliver stable performance in applications with dirty power or periodic power fluctuations. Most flame detectors require an external reflector to perform automatic visual integrity testing, while some apply a sapphire window to eliminate this external reflector, which reduces maintenance and faults related to a failed self-test from lens fouling. Flame detectors that come with a wide operating voltage range, a modular design for easy electronics replacement, field-selectable sensitivity and time delay, an automated visual integrity self-test and an available external test lamp are best suited for the challenges that are presented with optimizing a detector into its environment.

UV detectors look for the UV light emitted by a flame. A common operating range is 185 nm to 260 nm, which eliminates problems with sunlight and incandescent or fluorescent lighting. A drawback to UV sensors is that petroleum products tend to be opaque to UV wavelengths, which means that a UV sensor requires periodic cleaning to remove any oil that has accumulated on the lens.

Other materials that may affect sensitivity include smoke, dust, dirt, oil and grease, silicone based cleaners and standard window glass, all of which absorb UV energy. In addition, UV sensors are susceptible to false alarms caused by some types of lighting and by UV sources like arc welders operating in the vicinity.

UV/IR detectors combine a UV detector with an IR detector, and respond only if both UV emissions in the 185 nm to 260 nm range and IR emissions in the 4.4 µm range are present simultaneously. These ranges are sensitive to hydrocarbon and metal-based flame, while special tuning is required for hydrogen and silane fueled fires. UV/IR detectors offer good false alarm immunity and fast response, and are suitable for indoor and outdoor applications. Precautions in their use stem from the fact that some airborne contaminants absorb UV radiation, and that contaminants on the detector lens (steam, oil film or smoke) can reduce sensitivity.

Desirable features

 Traditional UV and IR sensors check the integrity of their optical paths by projecting a beam of UV or IR light to an external reflector (usually made of metal) that reflects the beam back to the detector. But this visual integrity test method has several limitations. The reflector can be degraded by accumulations of airborne contaminants (it can get dirty) or be corroded, and require periodic cleaning. It can be knocked out of alignment, creating a fault condition. The detector’s lens can also get dirty, reducing its sensitivity; in some cases, the lens can be coated with a material that blocks flame signals, but allows the test beam to pass through normally, giving a false impression of integrity. Fortunately there are IR sensors available with local emitters that can check sensor function without the use of an external reflector, which can save a substantial amount of maintenance.


 Hazard detection should be divided into three levels to provide the most comprehensive coverage. The choice of which technology to apply depends on individual applications and the conditions present at the facility. Multiple detection methods are often the best approach. When selecting a safety monitoring solution, it is a good practice to look closely at suppliers with a wide range of technologies and sensor types, and to work with a single and experienced source that can provide knowledgeable consultation on the best approach when considering all the options available, and ultimately will support the complete solution. HP

Have your say
  • All comments are subject to editorial review.
    All fields are compulsory.

Jaime Bárcena -México-

Well illustrated issuance about "3D" steps to save facilities against dangerous gas release. Such "3D" safety path must be completed with or within a hazop study (L, P, T, C) to define and also install safeguards against such risky problem. Overall, at the case, when such gas release is hydrogen leakage that is not easily detected , even when such gas is burning indeed.

Raymond P. Smith

A well written and very informative overview.

A couple of extra tips:-
High integrity ultrasonic gas detection in conjunction with point gas detection offers the prospect of operating traditional overhead deluge systems in modules to mitigate (reduce) the explosion overpressure per published data. High integrity ultrasonic currently being field trialed.

For triple IR flame detectors, (confirmed my manufacturer)utilising the two outer 'non overlapping' bands only means the detector can potentially be used as a stand alone double knock device.

Both the above comments potentially give you less bang for your buck.

T V Venkateswaran

Some details about PID VOC detectors, say for benzene, would have been useful


Very useful article

Related articles


Sign-up for the Free Daily HP Enewsletter!

Boxscore Database

A searchable database of project activity in the global hydrocarbon processing industry


Is 2016 the peak for US gasoline demand?




View previous results

Popular Searches

Please read our Term and Conditions and Privacy Policy before using the site. All material subject to strictly enforced copyright laws.
© 2016 Hydrocarbon Processing. © 2016 Gulf Publishing Company.