**1. Introduction**

Fire risk is the likelihood of a fire or explosion to occur, which is affected by any unique feature or potential hazard. Increases in indoor combustibles, diverse high-rise buildings, and population density have led to increased fire risks. Fire risk assessment is necessary for maintaining the safety of buildings and designing fire safety measures. Insufficient measures can result in massive casualties and damage to property. Risk assessment considers the structure of the building, its contents, layout, wind speed, and escape routes in the event of a fire.

Fire risk assessment can be of two types: quantitative and qualitative. Quantitative assessment methods are based on frequency analysis of entities that may cause fire and explosion and a consequence analysis of the potential damage of an accident to humans or property [1–3]. Examples of quantitative assessment methods include fault tree analysis (FTA), event tree analysis (ETA), cause–consequence analysis (CCA), layer of protection analysis, risk matrix, and the frequency–number of fatalities (F–N) curve. On the contrary, qualitative assessment methods comprehensively collect the opinions of many experts and include what-if analysis, hazard and operability, and process hazard review. In Korea, fire risk assessments are generally qualitative; a quantitative assessment method has not yet been universally accepted. Quantitative fire risk assessments are needed for liquefied natural gas plants, aerospace facilities, and other facilities with major fire hazards [4,5]. Fire risk assessment for existing nuclear plants and safety-critical systems can widely vary in approaches and outcomes [6].

Recently, in order to promote national economic growth and national interest, not only government space activities, but also commercial space activities of private companies

**Citation:** Kim, H.J.; Jang, K.M.; Yeo, I.S.; Oh, H.Y.; Kang, S.I.; Jung, E.S. Quantitative Risk Assessment for Aerospace Facility According to Windrose. *Energies* **2022**, *15*, 189. https://doi.org/10.3390/en15010189

Academic Editors: Roberto Alonso González Lezcano, Francesco Nocera and Rosa Giuseppina Caponetto

Received: 23 November 2021 Accepted: 24 December 2021 Published: 28 December 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

are considered as one of the key areas of activity [7]. According to the establishment and operation of space launch sites increases through the active space activities, various risks may occur in the aerospace facility, such as damage to structure and human life due to complexity and risk factors of aerospace systems. As such, a quantitative risk assessment of all parts, procedures, and operations of the aerospace facility is required [8–10]. Aerospace facilities perform various operations that pose a fire risk, such as rocket transfer and assembly, fuel and oxidant charging, and maintenance of electronics and general machinery. If a launch is rescheduled, the charged kerosene should be immediately extracted and then recharged. Due to the large amount of kerosene involved in the extraction and charging processes, as even a small leak can pose a fire risk. Jet and pool fires can be caused by fuel and oxidant leakage, tank failure, and catastrophic tank rupture. The potential of an explosion and secondary damage increases with the size of the leak. Thus, a quantitative fire risk assessment should be conducted to minimize the primary and secondary fire damage that could occur in aerospace facilities, with a special focus on the metal hose that is connected to the tank inside the rocket for supplying and extracting fuel [11,12]. Due to the metal hose posing a high risk in the event of a fuel leak, quantitative risk assessment should be performed for metal hose.

In this study, UK Health and Safety Executive (HSE) probability data was used to analyze metal hose failure and ignition frequency, and a consequence analysis was performed to analyze the fatality of jet fire. Wind rose was used to analyze the effect of wind direction and speed, which are the most important factors in the case of a jet fire. Finally, the individual risk assessment to the aerospace facility from metal hose jet fire was performed, considering the probability of failure and ignition, consequence of jet fire, wind direction and speed through comparison with the HSE risk standard.

#### **2. Methodology**

Validation for quantitative risk assessment can have several meanings that are disputable. As such, quantitative risk assessment of validation aspects, considering the objectives, expected results may be necessary [13,14]. Specifically, sampling proper data and its methods are most important to conduct quantitative risk assessment exactly. Therefore, in this study, through the validation approaches, a quantitative risk assessment was performed.

#### *2.1. Validation Approach*

Quantitative risk assessment is expressed as multiplying the frequency by consequence of a hazard factor leading to an injury or disease. In the case of frequency, it is the probability of occurrence of accidents and diseases for hazardous and dangerous work. In the case of consequence, it is the degree of exposure to human and material damage. As such, the quantitative fire risk assessment risk is a method of predicting the degree of risk for multiplying the frequency by consequence values of fire, according to the accident scenario, which are dependent on various parameters. Finally, after the quantitative risk assessment, it establishes mitigation to solve the workplace risk factors, such as installing emergency warning devices and securing a minimum safety distance [2,4,15]. Quantitative fire risk assessment consists of five methods, as shown in Figure 1 [16]. This section is introduced based on quantitative fire risk assessment to perform this research.

**Figure 1.** Quantitative fire risk assessment procedure.

#### 2.1.1. Part 1. Risk Assessment Objective (Select Accident Scenarios)

This part means to identify hazardous source and accident scenarios that match installation specific interests. In some cases, relevant guidelines and standards could be applied for this part. Selecting proper accident scenarios is important. When a risk is calculated by considering an accident with a very low frequencies and low impact, it takes a lot of time and money to control the risk. To select the proper hazardous source, a P&ID (Piping and Instrument Diagram) and aerospace facility drawing were analyzed, and a metal hose was selected as the most dangerous source.

#### 2.1.2. Part 2. Fire Risk Information (Collect Relevant Data Sources and Risk Calculation Method)

This part is necessary to conduct quantitative risk assessment adequately. A frequency data, which is needed to calculate fire occurrence probability, could be collected to analyze accident data from HSE, OGP (Oil and Gas Producers), NFPA (National Fire Protection Association), and HIAD (Hydrogen Incident Accident Database) [4,5]. A consequence could be calculated from various empirical equations or differential equations. Since solving differential equations is very difficult and complex, empirical equations based on various experiments are used to calculate the consequence and risk analysis. A fatality of jet fire and consequences are calculated based on the TNO Green book [17] and Han and Weng [18], considering mass release rate and probit function. An individual risk is calculated from fatality, frequency of jet fire, and weather conditions, which is made from HSE; a simplified approach to estimating individual risk [19]. A jet breakup length that is corresponded to velocity, density, surface tension, and viscosity, where the empirical equation was validated from Miesse [20]. Considering weather conditions in the Republic of Korea, wind rose data from Korea Meteorological Administration (KMA) was used.

#### 2.1.3. Part 3. Frequency (Selecting Proper Database and Applying)

As a result of the accident probability database analysis, the specific accident probability data for a metal hose does not exist. As such HSE probability database applicable to general leak situations were selected.

In this part, the metal hose fire probability was calculated from metal hose failure and ignition probability values using the HSE and cox et al. database [21,22].

#### 2.1.4. Part 4. Consequence

The consequence of accidents can be analyzed through fatality of overpressure, radiation intensity, and toxic gas concentration. In the case of jet fire, fatality is caused by the radiation of the flame. In this study, the fatality area and distance were calculated using the Han and Weng [18] equation based on mass release rate and probit function.

#### 2.1.5. Part 5. Quantitative Fire Risk Assessment Presentation

Quantitative fire risk assessment can derive the final individual risk using frequency and consequence analysis. In this study, wind rose data was applied to HSE's individual risk equation [19] to consider both wind direction and speed.

#### *2.2. Scenario and Boundary Definitions (Metal Hose)*

The target scenario was set in the space launch facility at Goheung, Republic of Korea, which was operated by the Korea Aerospace Research Institute. The risk assessment focused on the metal hose, which was highly likely to cause a jet fire due to leakage, failure to switch off, cracking, or excessive flow rate. The risk of a jet fire was calculated by considering the size of the metal hose, internal temperature and pressure, flow rate, and location. Since the metal hose charges and discharges fuel, it was always exposed to danger in the event of a fire or accident. If the metal hose and rocket were separated, a larger fire may occur. Figure 2 shows the schematic of the metal hose used at the Goheung rocket launch facility, respectively. Table 1 presents the properties of kerosene [23,24]. As shown in Figure 2, the metal hose could be divided into three stages. A jet fire may occur at any stage. However, jet fires at higher stages will not result in injury to humans; thus, the fire risk assessment focused on the lowest stage (i.e., metal hose 1), since this part was most likely to cause injuries to humans.

**Table 1.** Properties of kerosene [23,24].


**Figure 2.** Schematic of metal hoses connecting the rocket to the umbilical tower.

#### *2.3. Individual Risk*

Individual risk (IR) refers to the probability of personal injury in hazardous facilities when a fire occurs near an ignition source [19,25]. It can be expressed in two ways. The first is by plotting risk contours to designate the range around a hazardous facility for the same risk level. The second is a two-dimensional profile that indicates the risk level according to distance from the hazardous facility. IR can be calculated as follows [19]:

$$IR\_k = \theta\_k \cdot \sum\_{i} p\_{\text{loc},i,k} \cdot FoF\_i \tag{1}$$

where *IRk* is IR of population group *k*; *θ<sup>k</sup>* is the overall fraction of time that population group *k* is in the area; *ploc*,*i*,*<sup>k</sup>* is the probability that population group *k* is at location *i*; *FoFi* is the frequency of fatalities at location *i* [19,26].

*FoFi* can be calculated as follows [19]:

$$FoF\_i = \sum\_j f\_{co,j} \cdot p\_{fat,i,j} \cdot p\_{wother,j} \cdot p\_{direction,i,j} \tag{2}$$

where *feo*,*<sup>j</sup>* is the frequency of event outcome *j*; *pf at*,*i*,*<sup>j</sup>* and *pweather*,*<sup>j</sup>* are the probabilities of fatality and weather conditions produced by event outcome *j* (from meteorological data, 1 for weather independent event outcomes), respectively; *pdirection*,*i*,*<sup>j</sup>* is the probability of the direction required to produce event outcome *i* (related to the wind rose and cloud width for gas dispersion events, one for omnidirectional events). For the sake of simplicity, the following assumptions are used to calculate the individual risk [19,27,28]:


#### *2.4. Jet Fire Fatality*

The fatality of a jet fire is calculated from the thermal radiation. In the case of the jet fire fatality probability function, the Equation (3) considering the thermal radiation exposure time to the target and the thermal radiation flux is used [17,29,30].

$$P\_r = -14.9 + 2.56 \ln \left( t \cdot I^{\frac{4}{5}} / 10^4 \right) \tag{3}$$

where *t* is the time that the target is exposed to radiation. The exposure time is generally assumed as 30 s in urban settings. *I* is the thermal radiation flux [30].

Second, if you replace the thermal radiation flux, you can use the Equation (4) using the leakage flow rate and the distance between the target and the center of the flame zone.

$$P\_r = 16.61 + 3.4 \ln \left( \frac{Q}{R^2} \right) \tag{4}$$

where *Q* is the mass flow rate of leakage; *R* is the distance between the target and the center of the flame zone [30].

In the study, the probabilities of 99%, 50%, and 1% fatality can be calculated according to the radiation, while the areas and distance for these fatalities, due to a metal hose failure, can be calculated using Equation (5) with the probability values of 7.33, 5, and 2.67, respectively [17,30].

$$r\_{jct,99} = 3.915\sqrt{Q},\ r\_{jct,50} = 5.514\sqrt{Q},\ r\_{jct.1} = 7.768\sqrt{Q} \tag{5}$$

The average fatality probability in each area can be calculated as follows [29,30].

$$\frac{\int\_{0}^{3.915} Pdr}{\int\_{0}^{3.915} dr} = 1, \frac{\int\_{3.915}^{5.514} Pdr}{\int\_{3.915}^{5.514} dr} = 0.805, \frac{\int\_{5.514}^{7.768} Pdr}{\int\_{5.514}^{7.768} dr} = 0.172 \tag{6}$$

Figure 3 shows a schematic of the three fatality areas around a jet fire.

**Figure 3.** Schematic of fatality probability according to areas around a jet fire point.

### *2.5. Kerosene Atomization*

During the charging process, fuel flows through the metal hose at a velocity of 4–100 m/s and a pressure of 1.5–10 bar. If the metal hose ruptures, the high velocity of the leaking fuel can cause a jet fire, where the fuel is atomized or evaporated as it combusts. At room temperature (15–20 ◦C), liquid propellants evaporate more slowly than liquefied propellants; thus, a pool fire is more likely to occur [31–34]. However, a jet fire can occur under high velocity and pressure conditions. Khan et al. [35] reported the combustion characteristics of kerosene droplet. Due to the research data, a kerosene droplet under the 125 μm, was immediately burned by ignitor. Therefore, atomized kerosene was sufficiently burned by ignitor, and jet fires can be generated. The size and shape of the atomization depend on *We* and *Re* of the leaking fuel. Liquid fuel frequently atomizes when the velocity was over 100 m/s [35–37]. An atomization characteristic was correlated with leakage velocity, and when the leakage velocity was increased over 100 m/s, mostly liquid fuels, which were passed atomization length (jet breakup length), were atomized and immediately vaporized, but liquefied fuels were atomized and evaporated under 100 m/s easily [31,32].

As shown in Figure 4, the atomization of the leak jet occurred at a constant breakup length. Miesse [20] had and established Equation (7) for calculating the atomization length. The Equation (7) was suitable for the density range 300 to 1500 kg/m3, and viscosity range 0.01 to 10 cp conditions. The position where kerosene atomizes can be calculated as follows [32]:

$$L\_c = 538 \, d \, \mathcal{W} \epsilon^{0.5} \, \mathcal{R} e^{-0.625} \tag{7}$$

$$\mathcal{W}\varepsilon = \frac{\rho V^2 d}{\sigma} \tag{8}$$

$$Re = \rho Vd / \mu \tag{9}$$

where *Lc* is the breakup length, *d* is the diameter, *We* is the Weber number, and *Re* is the Reynolds number. Table 2 presents the parameters of the metal hose according to its diameter. With the metal hose, it was impossible to meet the high-pressure conditions for kerosene atomization. Preferentially, a breakup length is calculated to consider fatality areas accurately, since the jet fire occurred after the breakup length. As show in Table 2, the minimum breakup length was 3.2 m. In this study, the accurate fatality area was described in Section 3, using the breakup length value.

**Figure 4.** Kerosene atomization.

