Selecting and Mapping Flame Detection

At Micropack we have been immersed in flame detection for decades. We understand flame detection from development of traditional and innovative detection technologies, to modelling flame detection coverage, to maintaining the devices while understanding their field behaviour, and of course through testing, testing and more testing.

From this experience we understand the challenges in selecting technology, and modelling detector capability and performance.

With so many factors to consider, flame detection can seem overwhelming. Considering factors including not only obstructions, but the configuration of such obstructions, relative distance between the flame and the obstructions to the device, sensitivity settings of the device, how marginal the detection would be based on the inverse square law, the environmental situation (i.e. sunlight flicker and intensity, humidity, fog, rain, background radiation from hot surfaces, flare radiation etc.). It is critical to remember a detector will perform differently in Indonesia compared to Kuwait, compared to the North Sea.

Detection technology is also a critical consideration. Even within the same overarching technology (i.e. multi-frequency IR), different devices within the same technology family show significant variations in detection performance. This is due to the specific wavelengths being analysed by each device, and how the onboard algorithms analyse the signals generated from the combination of the environment and the flame. Two devices using the same technology, for example, will demonstrate varying sensitivity to sunlight, background radiation, and demonstrate different sensitivities to different flame characteristics (i.e. one device may be sensitive to a methane plume, while the other is more sensitive to a methanol pool fire).

This takes us to the differing detection capability based on the flame being burned. The difference in spectral emission between a Methane, N-Heptane, LNG and Methanol fire vary to the human eye and often even more so in the radiant bands the flame detectors may be analysing.  

Another critical factor is the region of the flame being detected. In its most simple form, for example, certain devices will be more sensitive to the upper flickering portion of the flame and others to the base area of the flame, with obstruction geometry compounding this effect. Detectors do not solely rely on a total volume of radiant energy which is ‘visible’ around an obstruction.  

The orientation of the device is also critical. For example, optimal results in a test burn may be achieved by placing a detector at a low level looking perfectly horizontally at an elevated fire, but this is not reflective of true application and detection performance can therefore be misleading. Understanding these nuances is critical in detection placement. Detection performance can also be artificially improved during testing if the device is set to its highest sensitivity (which is not what would then be applied in the field due to excessive false alarms) and shielded from environmental factors like sunlight. Designers must always scrutinise fire testing to ensure the correct detection technology is selected for the application, and also to ensure accuracy is implemented in flame detection coverage modelling.

This article serves as an introduction to the challenges surrounding flame detection. These are often over simplified and not well documented. A detector does not simply alarm to the presence of a flame, and detection performance will certainly not be consistent from day to day and across all combustible materials. This means it is critical for the designer to have a true reflection of the nature of detection capability and congestion factors. This is achieved first and foremost from an awareness of how the flame detector operates, which may result in the understanding that measurement of plume radiation does not equate to a detection prediction. To paraphrase Albert Einstein – what is important may not be measurable, and what is measurable may not be important.

This understanding of such critical factors can be achieved through professional and thorough testing. It is fundamental to flame detection development, design and mapping. It’s what we do. It is also why we recently upgraded our testing facility at Micropack HQ. We have filled our test ground with cutting edge data capture and environmental monitoring capability with online cloud-based video monitoring and data collection, while also expanding the range of flame burns we can run. We have also expanded the environments we can simulate with cutting edge water mist and radiation generation systems.   

To discuss this subject further (or any other F&G related queries) speak to your local Micropack representative to learn more about flame detection, ask questions, query site issues or ask advice on any projects you may be working on which require assistance. If you are local to Aberdeen we’d love to see you and burn some fires. If you are not local, Micropack are now set up to serve North and South America, Europe, the Middle East, and South East Asia/ Australasia locally.

With everyone’s increased reliance on remote conferencing, we can ‘beam’ you directly into our test ground and burn some fires regardless of your location. You can also learn more about flame detection and all of the factors discussed here at www.micropackfireandgas.com. Also request some of the papers published on the topic at [email protected]. Let us know your specific query and we will dig out the literature on the topic and assist.

At Micropack we live flame detection and we want our customers to get involved with the subject which has kept us captivated for decades.