Fire Consequence Modelling – A comparison of the results from PHAST and KFX
Sharad Gupta, Woody Wong
IRESC
Abstract
Fire consequence modelling is routinely performed as part of safety studies for onshore and offshore installations in the oil and gas industry. While PHAST software is commonly used for this purpose, the use of KFX software has increased considerably in recent times to provide better accuracy in the results. KFX uses Computational Fluid Dynamics (CFD) techniques for modelling the flow field, flame front and thermal radiation levels leading to a higher degree of accuracy in the results but is more expensive in terms of the time and hardware resource requirement. PHAST on the other hand employs empirical formulae and curve fitting techniques for fire modelling resulting in shorter timescales at the cost of accuracy.

The objective of this paper is to study the difference between PHAST and KFX in jet fire modelling, evaluate the potential impact on safety studies, and identify a correlation between the results obtained from both these software. Such correlations could then be used to scale the results obtained from PHAST for any facility.

In the current work, horizontal releases of methane were modelled for four different hole sizes (10 mm, 25 mm, 50 mm and 75 mm) and four different pressures (10 barg, 30 barg, 60 barg and 90 barg). The jet fire consequence results obtained by PHAST and KFX were extracted for three different thermal radiation levels corresponding to the damage criteria suggested by CCPS and UK HSE namely, 37.5 kW/m2, 12.5 kW/m2 and 5 kW/m2. These thermal radiation contours obtained from both the software were compared in terms of the maximum downwind distance and crosswind distance. Some suggestions are made on how the PHAST results may be interpreted and used in the industry.