Kenexis is happy to announce the release of Effigy™ Version 4.1.0. The release of this version introduces the plume fire algorithm for fire geographic mapping. This blog describes the details of the plume fire model, how it differs from the point source model used by most fire and gas mapping software, and when the use of the plume fire model is appropriate.
Background
As most Effigy™ users are already aware that Effigy™ supports the import of 3D models for fire and gas mapping. This import functionality is an advancement over the traditional methods of developing simplistic 3D models from geometric primitives which can be a time-consuming task. While importing 3D models directly provides benefits in time-savings it also introduces some complexity when it comes to the fire models used to analyze detector coverage, particularly when using the geographic coverage methodology.
When working with a traditional “geometric primitive” 3D model, it is common practice to only include objects which are large enough to obstruct a fire of a threshold size of concern (e.g., a 1ft x 1ft hydrocarbon pool fire). As a result, small obstructions such as small-bore piping, instrument tubing, conduit, etc. are often not included in the 3D model. Excluding these objects is appropriate and does not result in an overestimation of the achieved coverage because of a conservative assumption which is inherent to the “point source fire model” used by most fire and gas mapping software, including Effigy™.
The Point Source Fire Model
In the point source fire algorithm, a fire is assumed to occur at a single point in space (an infinitely small fire, radiating infinitely hot). This assumption results in a conservative calculation of coverage because, as we all know, a fire has volume. Making the assumption that a fire occurs at a single point in space is conservative even for the smallest fires of concern. This is why exclusions for minor obstructions when using the point source model is not only standard practice, but is often required to achieve an acceptable degree of coverage.
Let’s look at an example. The images below show the field of view for a single optical flame detector. In both images, the field of view of the detector is shown by the yellow region. In the image to the left you’ll notice a small green obstruction running horizontally across the image. This obstruction represents a 3-inch diameter pipe.
In the image to left, the field of view of the detector appears uninterrupted because your perspective is from the exact location of the detector’s lens. However, when the image is rotated and the pipework is shown in wireframe, as shown on the right, you can clearly see that the point source coverage algorithm has predicted an area of uncovered space directly behind the pipe from the perspective of the detector. The point source model has predicted that a fire centered at any location along the red obstructed region caused by the pipe will be undetectable. However, we know that in reality fires are not infinitely small point sources; we know that fires have volume. Intuitively we know that a 3-inch diameter pipe will not significantly impede the ability for a detector to detect a 1 ft diameter pool fire, this is why exclusions of such object is common practice.
When working with 3D models, what you will find is that operating with the conservative assumptions of the point source algorithm can result in overly conservative results because of the cumulative effect of these small diameter obstructions. For this reason, it is often necessary to implement a coverage algorithm which accounts for the volume of a fire. This algorithm is referred to as the “plume fire model”.
The Plume Fire Model
The plume fire model can be enabled on the Effigy™ study settings page as shown below. In the plume model a fire is represented as a 3-dimensional cuboid with equal x and y dimensions. The size of the fire is defined by the user using the plume width and plume height parameters.
In the plume fire model, the fire is assumed to be emitting thermal radiation equally over its entire surface. When the fire geographic coverage algorithms are run, Effigy™ will calculate the fraction of the fire surface which is observable from the fire detector. In doing this, the fraction of the fires total radiant heat output which is viewable from the detector can be determined. This is shown in the below images where a grid has been superimposed onto the plume (far left). Adding a small diameter pipe between the detectors point of view and the plume will obstruct a fraction of the plume which is visible (middle). This fraction of the plume will then be ignored by the fire coverage algorithms. After this obstructed area is accounted for, Effigy™ will then determine whether a fire is detectable based on the remaining observable surfaces of the plume (far right).
Now let’s look at the plume fire algorithm in action using Effigy™.
Just as before, we’ve introduced a 3-inch diameter pipe into the field of view of an optical flame detector. However, now the area directly behind the pipe is predicted as an unobstructed region.
Why? This is because the volume of the fire is sufficient such that a fire centered at a location directly behind the pipe will result in a plume which is largely still visible to the detector. While to pipe may be obstructing 10%-15% of the fire surface, the detector is still receiving sufficient thermal radiation to cause the detector to reach an alarm state.
Real World Application
The following is a real-world example of 3D fire and gas mapping analysis where the differences between the point source and plume fire models are apparent. The following images are of a manifold deck on an offshore facility. As you can see in the 3D view below, there is a significant number of minor obstructions in the area being modeled.
The following images display a plot view of the achieved fire detector coverage utilizing both the point source model and the plume model. The top image displays the achieved coverage utilizing the point source model. The bottom image displays the results using the plume fire model with a 1ft x 1ft x 3 ft plume.
As usually, the areas in green are covered by 2 or more detectors, areas in yellow are covered by a single detector and areas in red are uncovered.
Conclusions
Use of the plume fire model is highly recommended when performing fire and gas mapping using imported 3D CAD models which contain a high degree of detail. The effect of changing between the point source model and plume fire model are most pronounced when working with 3D models where the field of view of detectors are highly obstructed by small diameter piping or other minor obstructions.
Use of the plume fire model is not recommended for fire and gas mapping studies where simplified 3D models are developed from the geometry primitives within Effigy™, unless those models are developed with a high degree of detail. In studies with simplified 3D models, use of the plume fire model could result in a non-conservative prediction of detector coverage.