For every possible route, there is a method of control. For instance, ingestion can be controlled by reinforcement of rules on eating, drinking, and smoking; inhalation can be controlled by ventilation, respirators, and personal protection equipment; and injection and dermal absorption by proper protective clothing. Moreover, the smaller the particle size, the more probable it is for the particle to translocate from any of the entry routes into the circulatory and lymphatic systems following to the organs and tissues. Some particles, depending on their size and composition, may produce irreversible damages to cells or organelle injury. Figure 2.14 shows an example of a cell compared with different sizes of particles.

Chemicals evolve along with technology and the use or production of toxic chemicals. Several of the toxic substances that present hazards for people and the environment are produced, and some of them are generated as byproducts of the process. For instance, carbon monoxide, produced by an incomplete combustion process, is an asphyxiant gas. In general, to achieve complete combustion is very difficult as there are many reactions that can be involved during the progress of the reaction. Another example is nitrogen oxides, which can be produced during welding processes. Nitrogen oxides present an irritant effect but on inhalation could be fatal.

As mentioned, chemicals evolve along with technology, and every year the number of chemicals produced increases. Studies performed by Langley (1978) identified that in the decade prior to 1978, a total of 4 million new chemicals were identified. From such a large quantity, it was estimated that the United Kingdom had 300–400 new chemicals per year, and the United States tends to duplicate such values.

Some of the problems related to exposure of toxic chemicals vary based on the time of exposure. For instance, noxious effects can result from continuous exposure to a substance that presents a level of toxicity that has not been appreciated. However, some toxic chemicals may cause cancer and malfunction of different human organs. In the following paragraphs, a more detailed explanation about different hazardous particles is given.

The human body needs several metals for proper functioning. However, it has been proved (Shah, 1998) that exposure to high levels of these metals can have an adverse health effect. In general, the smaller the particle, the higher are the chances for it to get into the organism. Therefore, activities such as dealing with or manufacturing nano- and micro-sized particles should be considered as a serious occupational hazard. In general, the inhalation of dusts is well known to have an adverse health effect. Moreover, the health effect they can cause is a function of the material itself, the exposure duration, and the dose. For instance, exposure to metal dust, such as platinum, nickel, or cobalt, can cause asthma, while the inhalation of other metallic dusts can lead to pulmonary fibrosis and lung cancer. In the following paragraphs, some examples of dusts and the effect of being exposed to them are discussed.

As mentioned earlier in this section, all chemicals can cause adverse health effects. However, for this to occur, exposure to the chemicals has to occur at a sufficiently high dose. If the dose is not sufficiently high, depending of the chemical, normally there is no adverse health effects. Therefore, for toxic substances, it is important to establish a threshold dose at which adverse effects become evident. As mentioned, not all particles produce adverse health effects; the toxicity depends on various factors, such as size, composition, crystallinity, and aggregation, among others. Moreover, the toxicity of particles to organisms depends on the organism genetic complement, which provides either ways to adapt or fight toxic substances. It has been found that the inhalation of certain nanoparticles is associated with certain diseases such as asthma, lung cancer, and Parkinson‘s and Alzheimer‘s diseases. Further, research has proved the presence of nanoparticles in the circulatory system in several cases of heart disease, arrhythmia, or arteriosclerosis.

There are several other methodologies for risk assessment in the federal government. Some of them are managing the process (NAS/NRC, 1983), assessment and management of chemical risks (Rodricks and Tardiff, 1984), toxicological risk assessment (Clayson et al., 1985), and the risk assessment of environmental hazards (Paustenbach, 1989; Bridges, 1985; Nabholz, 1991; Becking, 1995; Frederick, 1995; Mannan, 2012).

Furthermore, according to the National Academy of Sciences (NAS) report, risk assessment consists of four different stages (National Research Council, 1983; Cohrssen and Covello, 1999):

At this point, various issues regarding toxic chemicals have been addressed. For instance, dose–response and TRA have been studied briefly. It is clear that there is a need for establishing thresholds to avoid adverse health effects. Still, the level of exposure to be expected has to be defined. This is performed by using the suitable risk criteria.

Toxic risk assessment has been used to define hygiene standards, especially in the United States. For example, for inorganic arsenic, OSHA used a linear dose–response model. In such a case, OSHA has stated that for every 1000 workers over a working lifetime, there could be a maximum of eight deaths at a level of 10 mg/m³, 40 at 50 mg/m³, and 400 at 500 mg/m³. On the basis of this standard, the dose was reduced from 500 to 10 mg/m³. To justify this decision, OSHA stated (Mannan, 2012):

Threshold limit values (TLVs) refer to airborne concentrations that correspond to conditions under which no adverse health effect is expected during a worker's lifetime. It is assumed that the exposure is going to occur 8 h per day and 5 days per week (Crowl and Louvar, 2011).

There are three different types of TLVs. They are described in Table 2.3:

Table 2.3 describes exposure under working conditions, but exposures under emergency conditions are determined using probit equations. However, this approach is very limited, because data are available and published for a limited amount of chemicals. A secondary approach is to specify a toxic concentration criterion above which the interaction with the toxic substance can have adverse health effects. Some of these methods and criteria are described next (Crowl and Louvar, 2011):

To perform a hazard analysis is usually very challenging. The level of complexity for this type of analysis is subject to the level of detail the hazard analysis team wants to go into. For instance, if there are potential hazards that could affect a nearby population, such detail is essential. On the other hand, it is possible to disregard a scenario that will not cause any harm to any person. For example, the team is evaluating many processes and operations, but they realize that from 10 operations, only three have the potential to harm people. It may be sufficient to analyze those three scenarios, because the other seven scenarios do not have a chance to affect the nearby population. There are several types of hazard analysis. All of them are different and present a different procedure of how to assemble a team and how to analyze such hazards.

According to OSHA, one of the easiest and possibly most effective methods is the job hazard analysis (JHA), also known as job safety analysis (JSA). In general, a JHA focuses on job tasks as a way to identify hazards before they occur. At the same time, it finds a relationship between the worker, the tasks, the tools, and the work environment. The main purposes of the JSA are to identify hazards that are not controlled and to take appropriate measures to either prevent the hazard or reduce it to an acceptable risk level. Table 2.4, identifies the major type of hazards that can be identified on a JHA (OSHA, 2002).

There are two different types of releases. The first one is the continuous release, and the second one is the instantaneous release. In general, there is a criterion to determine what type of release is occurring. A continuous release is defined by the following relation:

It can be noticed that there is a gap between continuous and instantaneous release. In this gap, the release behaves neither as continuous nor as instantaneous; it is a transient release. The release thus can be described as quasi-continuous or quasi-instantaneous. In general, quasi-continuous release can be treated as follows: the total amount of substance released Q₀ over the duration of the release T₀ gives an effective release rate q₀. With this release rate, it is possible to obtain the concentration at a certain distance. If the release is treated as an instantaneous release, the only parameter that has to be used is the total amount of substance released Q₀ (Mannan, 2012).

Most of the commercial or publicly available models belong to one of these categories and can be found in Table 2.9 (Luketa-Hanlin, 2006).

From this, it has been shown that at a fixed point the concentration fluctuates considerably in a vapor cloud. Griffiths and Harper (1985) mentioned that these concentration fluctuations have influence on the effective toxic load (Wilson and Simms, 1985; Wilson, 1991; Blewitt et al., 1987; Davies, 1987, 1989) and propose some methods to treat this influence.

There are different methods to mitigate dispersion and reduce the size of the toxic cloud. For instance, the inclusion of barriers in the system may diminish the size of the cloud. In general, barriers are installed near the location of a possible leak. The purpose of the barriers is to contain as much gas as possible and make it more difficult for the gas to go around it and keep dispersing. Several types of barriers are used; some substances are regulated to be stored in vessels and those vessels, at the same time, must be in dikes.

From the discussion regarding the different types of models, it is clear that the only model capable of predicting how a gas is going to be dispersed in presence of obstacles, like in a facility, is the CFD model (Mannan, 2012). Therefore, when simulations are going to be performed for a certain facility, it is preferable to use CFD models. However, as CFD models demand a great amount of time, a balance between precision and estimate should be made. In other words, using CFD models for simple geometries can be a waste of time, and this type of scenarios could be studied using integral models. Figure 2.16 illustrates how different outcomes are obtained when there are obstacles included in the scenario and when the scenario does not have any obstruction.

As can be observed in Figure 2.16, the obstacles present in the scenario significantly reduce the downwind distance the cloud could reach. This occurs mainly because there is an increasing turbulent phenomenon around the obstacles, which changes the flow field around them. It can also be observed from this figure that the cloud finds its way out of the obstructed area; however, it takes more time for this to occur. Additionally, the cloud size becomes smaller.

As it has been stated by Busini et al. (2012), there is a relationship between the height and position of the wall with the cloud dimensions in the open field. However, one cannot construct an infinite height/length barrier to reduce the risk. Busini addressed this issue and determined that for an obstacle to influence the cloud dispersion the wall has to have an adequate width and the height of the wall should be the same as that of the cloud to induce variations in the maximum affected distance. For instance, if the cloud is higher than the barrier, the barrier will not present much of an obstruction to the cloud. Therefore, before installing a barrier, design studies should be performed to determine the appropriate height and width of the barrier for it to fulfill its purpose (Busini et al., to be submitted for publication).

In the past decades, water sprays have gained a lot of attention from both academia and industry. It has been proved to be an effective engineering measure to mitigate the hazards of a dispersing cloud. Some of the reasons for the success of water sprays are their availability, the simplicity of use, and their efficiency (Uznaski and Buchlin, 1998). According to Hald et al. (2005), water curtains can be classified as full cone, full square cone, hollow cone, and flat-fan. All of them present a different flow pattern. Now, it is important to study the key parameters that can make water curtain successful when mitigating gas dispersion. Initially, one must focus on the atmospheric conditions. For instance, the ambient temperature and pressure should allow the water to be in the liquid phase and avoid getting into the freezing zone. Another important factor is the wind direction.

Several studies have been conducted to understand how dispersion is affected by water curtain. However, most of the studies have been performed with flammable gases such as LNG. Kim et al. (2013) discussed that there is less risk for the population and a couple of them have been performed with highly toxic substances such as hydrogen fluoride (HF) (Blewitt et al., 1987). In the first set of studies just mentioned, they concluded that water curtain has been proved to be effective in mitigating flammable vapor concentration below the flammability limit. In other words, the use of water curtain can reduce the dimensions of a flammable cloud. Now, if the same principle is applied to toxic substances, it is possible to reduce the size of the vapor cloud as well. Blewitt et al. (1987) reached the same conclusions in their experiments as well. They measured the effectiveness of the water curtain by its ability to make the HF dispersion end up near the source. For this system to operate efficiently, huge amounts of water should be released to the environment so as to capture more acid. Cornwell et al. (1998) performed a comparative quantitative analysis in evaluating different mitigation systems for HF. In this study, advantages and disadvantages of water spray were described. Table 2.5 presents such advantages and disadvantages.

Another important feature of this system of mitigation is that if a substance reacts with water, different chemicals could be used. Further, the effectiveness of water curtain depends on the following parameters:

In Table 2.6, a study performed by Cornwell et al. shows different scenarios where a water curtain was used to prevent the dispersion of HF to reach high distances.

Currently, there are several efforts to model how the dispersion phenomenon behaves with spray curtains. Again, in this case, the most desirable model to use is CFD as most of them have the ability to add water curtains in the system. An example of CFD models that could be used are FLACS and FLUENT. As previously discussed, the heavy gas dispersion models available can now treat situations where migratory features exist. Since the typical release occurs in a work environment, the gas dispersion will be strongly affected by the presence of buildings, which tend to enhance the dispersion. An illustration of a hazard assessment of a toxic release in the presence of buildings is given elsewhere (Deaves, 1987).

Another way to mitigate a toxic release is to find shelter. However, this mitigation measure should be used with caution. As has been described by Mannan (2012), once a human being starts performing physical activity, larger volumes of air are inhaled. Consequently, if there is a toxic substance mixed within the air, greater amounts of such substance are going to be inhaled and injuries could be reached much faster. Table 2.7 shows how the inhaled air increases with physical activity.

Moreover, as can be observed from Table 2.7, as the activity requires more effort, more oxygen is needed by the body. Therefore, before designating buildings for shelter, careful considerations should be made in designing the routes that could lead people to shelter and installing adequate ventilation systems. Finally, evacuation should be considered when very toxic substances have the potential to affect large populated areas with relatively small doses. Some substances that require evacuation if a release occurs are (Mannan, 2012):

The first type of scenario is the worst-case scenario. Here, all the contents of the vessel must be released to the atmosphere in 10 min (almost catastrophic failure). Moreover, the atmospheric conditions must be set to a relatively low wind speed (1.5 m/s) and a very stable Pasquill stability (F). Passive mitigation systems allow the dispersed cloud to concentrate near the facility and extend the hazardous distances (EPA, 1996).

The second type of scenario is a more realistic scenario, also known as the alternative release scenario. Credit can be taken for both active and passive mitigation systems when developing the alternative release scenario. One must address the key parameters that may affect the dispersion of the toxic substances. For instance, consider the situation where a gas is being discharged from an orifice in the vessel. In this case, if we have a bigger hole, the discharge rate is going to be relatively bigger than if we have a smaller hole. Therefore, it is important to be aware of the selection of the orifice size prior to setting up any type of calculations. The following are the most important parameters the reader should be aware of (Crowl and Louvar, 2011):

In general, when performing facility siting studies, the area of the largest pipe connected to the vessel is considered. Still, this will give overestimates of the consequences; however, it will be a conservative estimate. Table 2.8 illustrates typical guidelines for the selection of process incidents.