Selecting a basis of design for combustible gas detection involves determination of the threshold volume of combustible gas, above the lower flammable limit (LFL), which the gas detection system is intended to detect.  This threshold volume is referred to as the “design basis” cloud size.

Determination of a design basis cloud is necessary in gas detector mapping because it is impractical to expect any well designed fixed gas detection system to detect a release of any size.  Small, fugitive releases occur often in a typical chemical process environment due to maintenance and operations activities (e.g. bleeding impulse lines, sampling process streams, etc.).  It is impractical, and undesirable, for a fixed gas detection system to detect these releases for two reasons:

  1. Many of the small, expected, releases which occur as part of normal operations do not pose a significant hazard with respect to fire. Small volumes of combustible gas, if they were to ignite, would not be expected to result in a fire with any measurable consequence to personnel or assets.  Therefore detection of these releases does not provide any measurable benefit with respect to reducing the risk of a fire.
  2. Detection of extremely small volumes of combustible gas would require a large number of detectors located in extremely close proximity to potential leak sources. This would result in a system which is prohibitively expensive to design, build and maintain.

The consequence of a combustible vapor cloud fire is measurable by one of two parameters; thermal radiation and overpressure effects.  When a combustible vapor cloud ignites a flame front is formed.  As the flame consumes the combustible vapor energy is generated, which accelerates the flame front.  While all fires will generate thermal radiation significant damaging overpressure effects only occur at high flame velocities.  Therefore two types of vapor cloud fires are classified; fires with low flame velocities (flash fires) and fires with high flame velocities (vapor cloud explosions).

A fire with a low velocity flame front is referred to as a vapor cloud flash fire.  The consequence of a flash fire is characterizes by the size of the fire and the thermal radiation emitted by the flame.  The hazardous area impacted by the fire is typically limited to the area where combustible gas, above its LFL, is present before ignition.  Flash fires are short-lived transient fires and are typically survivable by personnel exposed to the flame if appropriate personal protective equipment (PPE) is worn.

A fire with a high velocity flame front is referred to a vapor cloud explosion, or VCE.  The consequence of a VCE is characterizes by thermal radiation as well as the magnitude of the overpressure generated by the fast moving flame front.  Based on research done of the UK HSE, flame front velocities greater than 100 m/s are capable of generating overpressure effects which exceed 150 mBar (See Blog from Edward Marszal – Origins of the Five Meter Grid for Gas Detector Spacing).  Overpressure in excess of 150 mBar is sufficient to cause damage to non-reinforced concrete structures as well as producing thrown debris.  The hazardous area impacted by a VCE is not limited to the area where the explosion occurs.  Personnel outside of flammable cloud have the potential to be impacted by thrown debris and/or failed structures.  The consequences of a vapor cloud explosion on personnel are often fatal and are much more severe than flash fires.

Typically, the design basis cloud size is selected on the grounds of reducing the risk of a Vapor Cloud Explosion (VCE).  Therefore, the intent of the gas detection system is to be capable of detecting any flammable vapor cloud with sufficient volume to generate a VCE in order to limit the duration/magnitude of the release and reduce the potential for ignition.

The volume of a flammable cloud capable of producing a VCE is a function of the environment in which the cloud is located.  Confinement (e.g. blast walls, decks, etc.) prevent the flame front from expanding in one or more dimensions.  The result is an increase in the flame front acceleration in the remaining, unconfined dimensions.  Therefore a smaller volume of combustible gas is required to produce explosive overpressure in areas with a high degree of confinement.  Additionally, physical obstruction in the area of a vapor cloud fire create a turbulent mixing effect at the flame front, leading to increased mixing of fuel and air resulting in an increased rate of burning and therefore increased flame front velocities.  As a result, areas with a high degree of congestion are also more prone to a vapor cloud explosion.

Several studies have been done in industry to investigate the effects of confinement and congestion on flame velocity.  Research done by the UK HSE (ref 2) suggests that for an area with two degrees of confinement (flame expansion limited to one dimension) and a high degree of congestion (blockage ratio greater than 0.3), a flammable cloud which is 5 meters in diameter has the potential to result in explosive overpressure.  Therefore, for areas with a very high degree of confinement and congestion a design basis cloud of 5 meters should typically be used.

Additionally, studies have been performed to predict explosive overpressure effects when a combustible gas cloud is ignited in a typical offshore process environment (ref 1).  Based on the results of these experiments, an 8 meter diameter cloud was found to have the potential to result in explosive overpressure for areas with typical confinement (flame expansion in two dimensions) and a moderate degree of congestion (blockage ratio <= 0.3).  Therefore, it is suitable to select a design basis cloud size of 8 meter for areas which are characterized as having 2D confinement and moderate congestion.

For large uncongested, unconfined areas the most probable outcome of a vapor cloud fire, regardless of the size of the cloud, is a flash fire.  Where combustible gas detectors is required in these areas, a design basis gas cloud of 10 meters in diameter or greater can be used subject to engineering heuristics and expert judgment.

  1. Hjertager, B., Fuhre, K., Bjorkhaug, M., 1988, Gas explosion experiments in 1:33 and 1:5 scale offshore separator and compressor modules using stoichiometric homogeneous fuel/air clouds, J. Loss Prevention in Process Industries, vol. 1, pp. 197-205
  1. Health & Safety Executive (UK), 1993. Offshore gas detector siting criterion investigation of detector spacing HSE OTO 93/002,