Consequences Modeling of the Akçagaz Accident through Land Use Planning (LUP) Approach

In the study, consequences analysis of Akçagaz LPG Facilities accident was conducted.  The consequences analysis, modeling studies were performed by the use of EFFECTS 10.0 Software over two liquefied gas LOC (Loss of Containment) scenarios. One of the scenarios was G1: Instantaneous release corresponding to BLEVE (Boiling Liquid Expanding Vapor Explosion) and the other was G2: Release in 10 min corresponding to UVCE (Unconfined Vapor Cloud Explosion). The highest threat zone distance (1kW/m2 heat radiation distance) was determined as 1699 m, the lethal burn distance as 377 m and distance from the center cloud to threshold overpressure as 342.46 m with the G1 scenario. French, Italian and Austrian methodologies relating to LUP (Land Use Planning) context of the Seveso Directive, which was not implemented in Turkish legislation, were evaluated for BLEVE of The Akçagaz Accident. Three different modeling approaches for BLEVE including static, dynamic and rupture of the vessel were used and the results were compared to the LUP methodology. The value (height of the fire ball: 273m) closed to the actual accident situation (height of the fire ball:200-300m) was obtained with the use of the static modeling approach. The distance access to fragments of the tank was calculated as 409 m with the use of rupture of vessel modeling approach which was compatible with the actual accident value (~500m). High lethality, the beginning of lethality, irreversible effects, indirect/reversible effects radius of The Akçagaz Accident were calculated for each country LUP methodology. The determined distances with the use of static BLEVE model correlation were obtained at the highest value again. High lethality radius was determined for French and Italian as 173.37 m and 86.13 m, respectively. The LUP methodology used in France is said to be more restrictive based on the large impact distances. On the other hand, when the TOTAL specifications (GS EP SAF 253& 262) are considered, which are dependent on demand but very important, the threshold values for health effects are seen to be much more stringent.