(104bm) Design Optimization of Storage Facilities Taking Into Account the Domino Effect

Storing hazardous materials is a necessary but risky process. Historical analysis [1] reveals that 17% of the major accidents in the chemical industry happen during the storage process, and the National Fire Protection Association (NFPA) reported [2] that in 2009, 13% of the major fire accidents that occurred in the USA happened in storage facilities, causing losses of $69,980,000. These numbers demonstrate that it is necessary to continue working on the improvement of safety in dangerous substance storage facilities.

When a process unit suffers an accident, the effects (mechanical or thermal) this event can have on the surrounding equipment can trigger subsequent waves of accidents, which can increase the consequences of the initial accident significantly, both in space and time, this phenomenon is referred to as the domino effect, and it has been formally defined as a cascade of events in which the consequences of previous accidents are increased both spatially and temporally by following ones, thus leading to a major accident [3]. A historical analysis [4] of 225 accidents that have involved domino effect, and that occurred after 1961, analyzing origin, causes, materials involved, effects, consequences and most frequent sequences of accident as the main factors in the study revealed that thirty five percent of the accidents studied occurred in storage areas, which makes these types of installations the most prone to suffering cascading accidents, and that eighty nine percent of the accidents involved flammable materials; LPG was found to be the substance involved in more events.

This work presents a methodology that allows obtaining the optimum number of tanks to be used in the design of a facility where explosive or flammable materials are stored, in order to minimize the risk associated to the facility; this is achieved by combining quantitative risk analysis and optimization techniques to calculate the risk associated to the facility, depending on the number of tanks used in the design, the mass and type of substance involved, and the frequencies and possible consequences of the accidents that can occur. The methodology is based on the fact that the consequences of an accident are directly proportional to the mass of hazardous substance involved, which means that dividing the quantity of substance in more tanks will make the risk associated to the facility decrease; however, as more units are built the risk of domino effect occurring will increase. Then, an optimum number of units to use in the design can be found, for which the risk is minimized, taking into account the amount of mass involved in the accident and its consequences on vulnerable elements, and also the possibility that domino effect can occur, and will become a severer risk as more tanks are used.

To find the optimum design, a model that calculates risk based on the LOCs (and accidents that derive from them) described in the Purple Book [5] was designed, monetizing the consequences of the possible accidents and multiplying these costs by their frequencies of occurrence, and later combining the risk derived from all the studied accidents to find the total risk associated to the facility. Domino effect is studied and included in the methodology using a threshold approach: if the effects of a previous accident on another tank are higher or equal to a certain value, then this second unit will suffer another loss of containment event, that will lead (with a certain probability) to another set of accidents, and an increase in risk.

The results obtained in the work demonstrate that the optimum number of units to store a hazardous material can be found, and that quantitative risk analysis can be paired with mathematical optimization to form a powerful design tool for storage facilities. It also helps proving the fact that performing risk analysis in the initial stages of a project can help saving lives and resources, and that it is necessary to integrate risk analysis into all the design stages of engineering projects.

References:

[1] J. Casal, J. Vílchez, El Riesgo Químico y el Territorio, Revista Catalana de Seguretat Pública, November 2010.

[2] S. G. Badger, Large-loss fires in the United States-2009, Fire Analysis and Research Division, National Fire Protection Association (USA), November 2010.

[3] C. Delvosalle. Domino Effects Phenomena: Definition, Overview and Classification. First European Seminar on Domino Effects. Leuven, 1996.

[4] Darbra, R. M., Palacios, A., Casal, J. (2010). Domino effect in chemical accidents: Main features and accident sequences. Journal of Hazardous Materials: 183, 565-573.

[5] CPR 18E. Guideline for quantitative risk assessment (Purple Book). Directorate General for Social Affairs and Employement, TNO, The Hague (Netherland), 2005.