F-Guard IR3

Four Variants, Two Decisions

Two questions decide it. Is there exhaust in the detector’s field of view, which selects standard or CO2L. And do you want to see the scene as well as be told about it, which selects HD.

Eighty Metres

Detection reaches 80 metres for a 0.1 square metre n-heptane pan fire, the longest range in the F-Guard family.

Range decides detector count, and detector count decides installed cost. Across a large open process area, a tank farm or a loading facility, the difference between a 40 metre detector and an 80 metre one is often half the detectors, half the cabling and half the commissioning.

At shorter distances, the detector is quicker. The HD and CO2L variants detect a standard pan fire in 1.5 seconds at up to 30 metres, and the HD in 1.2 seconds at 15 metres. The standard IR3 responds in 2.6 seconds at up to 30 metres.

Forty Milliseconds

Fires and explosions are different events on different timescales, and a detector built for one will be wrong for the other.

A slow growing fire develops over seconds, and a detector responding in a second or two is more than enough. A vapour cloud ignition doesn’t develop. It just happens. The F-Guard IR3 has an ultra fast detection mode that responds within 40 milliseconds to a fireball or explosion.

That speed is what lets the detector sit in front of a suppression system that has to act before a fireball spreads, rather than one that records the aftermath.

Why the CO2L Variant Exists

This is the most useful thing on the page, because it explains a problem people usually meet after installation rather than before.

Triple infrared detection works by watching the infrared band given off by hot carbon dioxide, the product of hydrocarbon combustion. That signature is strong and specific, which is what gives IR3 its range and its immunity.

The difficulty is that some environments produce hot carbon dioxide as a matter of routine: a jet engine running in an aircraft hangar, a helicopter on a pad with exhaust and downwash, diesel trucks idling at a loading rack. In those places, the very signature the detector looks for is being produced by normal operations.

The CO2L variant is built for exactly that. It gives fast, robust detection of heavy hydrocarbon fires in locations where combustion exhaust gases may be present, telling a fire apart from the exhaust around it.

If your protected area has running engines, specify CO2L. If it doesn’t, the standard variant gives you the performance without the optimisation you don’t need.

What the HD Video Changes

The HD variants provide high definition colour video of the monitored area, with clear imaging of a fire and of personnel at up to 30 metres.

The value is in what happens after the alarm. A responder can see the situation before entering the area: whether the fire is where the detector says, whether anyone is inside, and how far it has developed. Those are decisions currently made by walking in and looking.

Recording is automatic, from one minute before the alarm to up to three minutes after, with video and event data written to non volatile memory. For an incident investigation, that’s the difference between reconstructing events from a log and watching them. Video integrates over ONVIF Profile S on a standard IP network.

Three Wavelengths, One Field of View

The detector uses three infrared wavelengths in the 4.0 to 5.0 micron range, with clear separation between them.

One design detail is worth drawing out. All three sensors share the same field of view, 90° horizontal and 75° vertical, so they compare the same scene rather than three slightly different ones. Where sensors see marginally different areas, an object at the edge of one field can produce a mismatch that looks like a fire signature. Aligning the fields removes that source of false alarm.

Combined with flicker frequency analysis, that’s what lets the detector work in all weather and light conditions with high immunity to false alarms. Five sensitivity ranges, from Extreme to Very Low, let it be matched to the site.

Event Logging and Self Test

Alarms, faults and other relevant events are written to non volatile memory, so the detector holds its own history rather than relying on the panel to log it. On the HD variants, video is stored alongside.

The built in test runs automatically and can be triggered manually, checking window cleanliness and the overall operation of the detector. A dirty optics warning flags when the window needs cleaning, before it affects detection.

It’s worth understanding the limit, which applies to all flame detectors. The self test confirms the detector can see. It doesn’t activate the alarm outputs, so proving the relays, the loop, the panel and any suppression release needs an external flame simulator. Gastech supplies F-Guard simulators matched to each detector type.

Integration

Outputs are universal, covering 3 and 4 wire configurations, 0-20 mA in sink or source with HART 7, and fire, auxiliary and fault relays rated 2 A at 30 VDC. An RS-485 port provides Modbus RTU. An alarm delay of 0 to 30 seconds is configurable.

That range of options means the detector drops into an existing fire and gas system, rather than needing one built around it.

Built for Harsh Sites

The housing is 316 stainless steel, rated IP66 and IP67 (IP66 and IP68 on the CO2L), and NEMA 4X and 6P. It operates from -55°C to +75°C, with a high temperature option to +85°C on the IR3 and IR3-HD. All variants carry a 5 year warranty.

Where It’s Used

Refineries and petrochemical plant. Oil and gas processing and terminals. LNG facilities. Large open process areas where detection range decides detector count. Tank farms and bunded areas. Offshore platforms and vessels. Aircraft hangars and helipads, using the CO2L variant. Truck loading racks and fuel dispensing, using the CO2L variant. Turbine halls. Compressor stations. Mine site fuel storage. Recycling and waste handling facilities.