Journal of Loss Prevention in the Process Industries
Worst-case identification of gas dispersion for gas detector mapping using dispersion modeling
Introduction
Vessel leakage and rupture are the main sources of catastrophic events in the process industries, causing material dispersion, fire and explosion. Flixborough of England in 1974 (Report of the Court of Inquiry, 1975) and toxic gas release in Seveso (Italy) in 1978 (Assael & Kakosimos, 2010) are examples of events that occurred as a result of material release from vessels. The consequences of these events, in contrast to those that occurred before 1960, extended beyond the factory. Additionally, in Bhopal (India), one of the worst catastrophes in the process industries occurred for the same reason as a catastrophe in 1984 (Assael & Kakosimos, 2010). These events make material release awareness more necessary than ever. Therefore, gas detectors as tools to monitor alarming gas leakage are used in the process industries. One of the gas detector properties that affects its performance is its installation location. Thus, the most appropriate location must be chosen. In this study, a scheme for gas detector placement has been presented to optimize detector installation and increase the performance of gas detectors.
Currently, a high volume of chemicals with specific properties exists in the process industries. Thus, it is necessary to be able to predict the behavior of a released fluid to estimate potential injuries, make a rapid reaction plan and mitigate the consequences.
PHAST Software is a tool used for behavior prediction and dispersion modeling, presented by DNV in 1999. PHAST not only includes different event models but also makes the study of material release consequences possible, from leakage to explosion (Ruiz-Sánchez, Nelson, François, Cruz-Gómez, & Mendoza, 2012). PHAST is useful because it does not require large amounts of input data and its calculation time is very short. Additionally, it provides required data for the risk assessment of equipment and processes (Witlox, Stene, Harper, & Nilsen, 2011). Using this software, events are modeled based on process conditions such as pressure, temperature, material composition, material flow rate, atmospheric conditions and the properties of the material release point (Witlox, Harper, & Oke, 2009). By modeling the release consequences using PHAST software, the manner of dispersion and the amount of released material versus time can be determined. PHAST software graphs are presented as concentration versus distance and time. By looking at the results of the gas dispersion modeling at a specific time after release, optimized gas detector placement will be possible. The specific time is determined based on the Emergency Shutdown System of the unit and personnel ability during emergency response. This study discusses gas detector placement by dispersion modeling using PHAST 6.54 software based on the appropriate conditions.
Section snippets
Process conditions
Gas extracted from the earth must be properly treated for industrial and domestic use. One of the treatment steps is to separate extra water and hydrocarbons. The Gonbadli Dehydration Unit of the Khangiran Gas Refinery was designed to control the dew point of gas. The gas first passes through a loop. After the heavy hydrocarbons are separated, water and the remaining hydrocarbons enter the Gonbadli unit. 99.8% of the feed gas is methane (based on the process description of the Gonbadli
The worst-case selection
In PHAST, events are modeled based on the process conditions, atmospheric conditions and release point properties. To increase the accuracy, dispersion should be modeled based on the appropriate input data. However, it is difficult to examine the most appropriate weather conditions, release elevation and release point on the plot plan. The most appropriate conditions are represented by the situation where the release detection will be definite at a desirable time in the other conditions as
Gas detector placement
The worst-case scenario for gas dispersion was determined in the previous section. The selected conditions are appropriate for dispersion modeling of the other vessels. In this paper, the detectors were located for the absorber tower V-1600 as an example. With this in mind, gas dispersion versus time graphs were used to determine the optimal detector location. Based on the worst-case studies, dispersion was modeled for gas release from the bottom of the vessel with a height of 1.65 m in the
Conclusions and recommendations
Release in vessels can cause material dispersion, fire and explosion in the process industries (Fig. 11). Modeling of this dispersion is useful to predict the behavior of the released material and to properly place gas detectors. However, weather conditions, release elevation and release location on the plot plan, which are required for modeling, are variable. Thus, dispersion should be modeled based on the most appropriate cases (the worst-case scenario for gas dispersion). In this paper, gas
References (7)
- et al.
Application of the accident consequence analysis in the emergency system design of an SI cycle hydrogen production plant
International Journal of Hydrogen Energy
(2012) - et al.
Modeling of discharge and atmospheric dispersion for carbon dioxide releases
Journal of Loss Prevention in the Process Industries
(2009) - et al.
Modeling of discharge and atmospheric dispersion for carbon dioxide releases including sensitivity analysis for wide range of scenarios
Energy Procedia
(2011)
Cited by (19)
Optimization of gas detector placement considering scenario probability and detector reliability in oil refinery installation
2020, Journal of Loss Prevention in the Process IndustriesCitation Excerpt :These accidents demonstrate besides good management of process safety and safety culture,in-time detection of hazardous gas release is a key element to prevent major accidents in process plants. However, one of the challenges to ensure efficient detection is the proper placement of gas detectors (Legg et al., 2012; Meysami et al., 2013; Rad and Rashtchian, 2016). Traditional methods applied for gas detector placement are experiential, typically based on standards, regulations and recommended practices, such as API (API, 2001), ISA (ISA, 2003), and so forth (IEC, 2007; ISA, 2010; NFPA, 2007).
Designing an effective mitigation system based on the physical barrier for hazardous chemical leakage accidents
2019, Journal of Industrial and Engineering ChemistryLeakage risk quantitative calculation model and its application for anaerobic reactor
2017, Journal of the Taiwan Institute of Chemical EngineersCitation Excerpt :According to the above-mentioned three meteorological groups, as well as the combustion limit (UFL, LFL and LFL Frac), the radiation intensity of jet fire (4 kw/m2, 12.5 kw/m2 and 37.5 kw/m2), explosion overpressure range (0.02 bar, 0.13 bar and 0.2 bar) and other five target parameters in the calculation results, we are able to obtain the fitting formulas of all dependent variables respectively, and then take these formulas as the mathematical model used for later analysis and forecast. According to the above results gained from the orthogonal experiment, we can know that the working condition of the worst case scenario (WCS) is the leakage of ground broken pipe, namely, the leakage caliber is 150 mm and the leakage position is on the ground [18]. According to the size and position of typical leakage hole of API 581 [19], some potential leakage scenarios are classified based on actual situation in Table 5.
Long-term consequence and vulnerability assessment of thermal radiation hazard from LNG explosive fireball in open space based on full-scale experiment and PHAST
2017, Journal of Loss Prevention in the Process IndustriesCitation Excerpt :The fatalities, which are the degree of harm to humans and surroundings, were applied based on the following criteria (Witlox et al., 2012; Woodward and Worthington, 1999; Crowl and Louvar, 2002b). According to the research results of thermal radiant flux criterion by China Academy of Safety Science and Technology (Witlox and Oke, 2008; Witlox, 2010b; Meysami et al., 2013), as presented in Table 2, these thermal radiant values vary from 1.6 to 37.5 kW/m2 and it is classified with five harmful level. As for the fatal zone, there is a 1% probability of death that if the person stays within the area for 10 s, comparatively, the people will not suffer from any harmful effects even if they have no protection in safe zone.
Utilization of regression technique to develop a predictive model for hazard radius from release of typical methane-rich natural gas
2016, Journal of Loss Prevention in the Process IndustriesCitation Excerpt :The resultant data formed a complete database (Table 3) which provides a regression-based model over it. Effective factors on hazard radius are in two categories: meteorological parameters and process parameters (Meysami et al., 2013). The area classification process is NOT a full hazard assessment which takes account of the range of meteorological and process conditions and release events (Roberts, 2001).