General procedure of QRA. 

Contexts in source publication

Context 1

... makes it possible to determine the acceptability of a process and route unacceptable processes to the risk reduction procedures. Generally, QRA is composed of six steps as shown in Figure 1. After the system and objective are defined, the major hazards are identified and their consequences and frequencies are analyzed. ...

Context 2

... VLLE and physical properties such as the mass density of the P_06 stream are calculated with the Peng−Robinson equation of state (EOS) using the Aspen HYSYS V.8.4. The results through process simulation are summarized in Figure 9 and Table 6, and those of the consequence analysis for the P_06_IF_10 mm scenario are plotted in Figure 10 and listed in Table 7. The flash fire effect is remarkably reduced at the new operating pressure of 0.52 barg due to a lower liquid flow rate and static inventory. ...

Context 3

... addition, the lower release rate of the 0.52 barg case due to the smaller pressure difference between the process stream and atmosphere decreases the ignition probability and explosion conditional probability at the ETA, as listed in Table 8. Based on the new frequency data calculated using the same method as in the original case listed in Table 3, the F−N curves for the societal risks of two pressure conditions are shown in Figure 11, which clearly shows that the risk is reduced to the level of the ALARP region. Moreover, the total risk integrals decrease from 1.41 × 10 −3 /year to 1.02 × 10 −3 /year, which is a more than 27% decrease. ...

Context 4

... order to analyze the sensitivity of the total risk to the development degree of the isolation process, a QRA simulation is recursively performed, increasing conditional probabilities of success for isolation up to 50% of that for the original case, with a step size of 10%. The total risk integrals and F−N curves for the societal risk are shown in Figures 12 and 13, respectively. The associated risk is drastically reduced when the probability of isolation success increases from 30% to 40%, which shows that it is desirable to develop the isolation process until the probability of success is improved to 40%. ...

Context 5

... working capital is considered to be 15% of the total permanent capital investment C TCI . Figure 12. Total risk integral with respect to degree of isolation success probability. ...

Context 6

... risk assessment (QRA) makes it possible to quantify the existing risks of an installation as data for determining whether they are acceptable, as well as assist in the prioritization of decisions to reduce unacceptable ones. 1 From the advent of discussions on the issues related to nuclear safety by Rasmussen in 1975, 2 extensive research to quantify the potential risks of chemical plants has been conducted for about 40 years. Although past and present studies have focused mainly on the general methodology for QRA integrated with loss-prevention technologies, 3 − 5 future research would concentrate further on the practicality, reliability, and accuracy of risk assessment methods. The majority of these studies consist of hazard identi fi cation integrated with the use of fuzzy relations or Bayesian networks in process facilities to re fl ect human error or domino e ff ects, 6 improved computational fl uid dynamics (CFD) calculations for modeling three-dimensional fi re and explosion or toxic releases scenarios, 7 − 10 and real-time risk analysis based on advanced process control. Among these emerging topics, however, there are relatively few studies on the methodology for integrating QRA with both process and accident simulations. Domenico et al. 11 imple- mented a process simulation for the risk assessment in an early design stage of a future plant using a limited amount of information, and delimited safe regions for the operation. Nam et al. 12 studied a methodology for evaluating the life cycle cost of an o ff shore natural gas liquefaction process in relation to the risk expenditure at the conceptual design stage. However, the former study just considered a simulation case study of the normal and worst-case operations, and the latter compared the existing two liquefaction processes from an economic point of view. Thus, the results could not provide fundamental guidelines for designing safer processes. From this standpoint, risk-based process safety management which integrates the QRA methodology with process simulation is very important for chemical plant safety in that it can not only evaluate the resilience of the process with respect to possible disturbances from the safety perspectives 13 but also consider risk-reduction steps involving the modi fi cation of the existing process design or the construction of inherently safe designs 14 − 18 for new processes. The resulting design would have safer structures for plant-wide unit operation and optimal operating conditions with better economic e ffi ciency and safety. In this study, with the goal of carrying out risk-based process safety management of an existing gas oil separation plant (GOSP), particularly the gas treating unit (GTU) among the various subsections of the GOSP process, a new QRA methodology for the feasible integration of process design (steady-state modeling and simulation) and accident modeling ( fi re and explosion, toxic dispersion simulation) is proposed using commercial simulators. This methodology attempts to assess the potential risk at the preliminary design stage and modify the process design of an existing process to reduce the risk with several design alternatives, in order to design a safer and more reliable plant. Section 2 deals with the QRA methodology for the targeting process through integration of process and accident simulation using the static inventory concept. Section 3 performs QRA and analyzes the results in order to determine the target accident scenarios of which the risk needs to be preferentially reduced. Finally, sections 4 and 5 discuss two risk reduction methods: one for modifying the design in a practical and reliable manner, and the other for analyzing the sensitivity of the total risk to the degree of development for the isolation process. General Procedure of QRA. The potential risks of a chemical plant can be quanti fi ed through QRA. This makes it possible to determine the acceptability of a process and route unacceptable processes to the risk reduction procedures. Generally, QRA is composed of six steps as shown in Figure 1. After the system and objective are de fi ned, the major hazards are identi fi ed and their consequences and frequencies are analyzed. Usually, the frequencies are obtained from historical data such as the UK ’ s Health and Safety Executive (HSE) or Oil and Gas Producers (OGP). The total risk is assessed quantitatively considering both the consequences and frequencies of major hazards. If the overall risk level is judged to be tolerable, the QRA process ends and that of the system is ready to be managed. Otherwise, the process should go through the risk reduction step and de fi ne the system recursively (feed-back loop). There are two standards for the tolerability criteria of associated risk: Individual Risk (IR) and Societal Risk (SR). IR is the annual risk of death or serious injury to which a person at the speci fi ed position is exposed. Thus, it cannot re fl ect the overall risk of the plant with various population density. In contrast, SR considers the population density so that it is used as a measure of the risk to a group exposed to the e ff ects of an accident. In this study, SR is used as the standard for tolerability criterion as it can more reliably re fl ect the real plant safety. SR is expressed in terms of the frequency distribution of multiple casualty events, which is called an F − N curve. This F − N curve is divided into three sectors: “ Negligible, ” “ As Low As Reasonably Practicable (ALARP), ” and “ Intolerable. ” Accord- ing to HSE ’ s publication, 20 an individual risk of less than 1.0 × 10 − 6 (per year) is considered negligible, while it is intolerable when it is above 1.0 × 10 − 4 (per year) for the public and above 1.0 × 10 − 3 (per year) for workers. As the number of people a ff ected by an accident increases, its fatality probability unconditionally decreases. The ALARP region is located between these two areas, negligible and intolerable, and the total risk curve of a process should belong to the ALARP or tolerable region in order to be acceptable. GTU Process of GOSP. 21 The target process of this paper is a Gas Treatment Unit (GTU) of a Gas Oil Separation Plant (GOSP) with a target production rate of 230 kbopd, and a block fl ow diagram is shown in Figure 2. The well fl uid from the gathering system station enters the fi rst stage separator, in which the free gas and water are separated from the crude oil. The oil is heated in the crude heater and depressurized prior to entering the second stage separator for further separation. Approximately 44.9 MMSCFD of the free gas from the two- stage separator is routed to the GTU, and the remainder heads for the Integrated Gas Process Plant (IGPP), where power generation, dry gas production, and LPG production are operated. In the GTU process shown in Figure 3, the combined feed gas is routed through a pipeline (P_01) at 13.4 barg to the high-pressure (HP) fuel gas suction scrubber (V_01, V_03) for liquid knockout prior to being compressed to 35.8 barg (P_02, P_03) by the HP fuel gas compressor. Compressed gas then enters the HP fuel gas discharge cooler (air cooling) to cool down to 60 C and is further chilled to 35 C in the chiller package. The refrigerated gas is then routed to the HP fuel gas discharge scrubber (V_02, V_04) for heavy hydrocarbon and water knockout. The liquid (P_05 (1) and (2)) is separated into hydrocarbon liquid, water, and vapor in the three-phase separator (V_05), and the hydrocarbon liquid (P_06 (1) and (2)) is fed back to the upper process at 5.5 barg. Fuel gas from the discharge scrubber is superheated to 80 ° C by the HP fuel gas discharge superheater and fi nally sent to the power plant. A model of the GTU is simulated using Aspen HYSYS V8.4 based on the Process Flow Diagram (PFD). The plant layout of the overall GOSP and GTU is softly depicted in Figure 4 for reasons of con fi dentiality, and part counts of the associated equipment belonging to each pipeline or vessel (P_01 − 06, V_01 − 05) are based on the Piping and Instrument Diagram (P&ID). Inventory. The risk calculation is basically based on the static storage conditions (pressure, temperature, mixture composi- tion/property, and volume/mass inventory) of each piece of equipment such as vessels or pipelines. All of these conditions are the same for the process simulation, except the static volume/mass inventory, because the chemical process is basically based on the nonstatic model using the fl ow rate of the fl uid rather than the static inventory. This di ff erence makes it necessary to convert the fl ow rate (kg/h) of the fl uid to the static inventory (kg) in the equipment in order to perform QRA integrated with process simulation. Some additional data such as equipment size, fl uid mass density, and release duration in case of a release scenario are necessary to evaluate the static inventory from the fl ow rate calculated from process simulation. In particular, the release duration of a fl owing fl uid is determined mainly by an ...