The approximate lower explosion limits of certain gases in air are shown in Table 15.1. There is also an explosion upper limit, which is dependent on the oxygen content – of course a constant for ordinary air. However, the limits of explosion are dependent on the total pressure (Holtappels et al., 2001).

There are mathematical models in order to calculate the limits of explosion (Askar et al., 2008), for example, the Software GasEq^® and some additional software has been used to calculate the flammability limits of mixtures with, e.g., ethylene oxide, air, and inert gases at temperatures between 20°C and 100°C and pressures between 0.4 bar and 1.0 bar.

GasEq^® is a Microsoft Windows-based program with a Microsoft Excel interface. The program can be used to calculate the equilibrium of combustion. It is intended primarily for gas phase calculations, although there is some limited facility for condensed phases, such as soot (Morley, 2005).

An ideal odorant should have the following physical and functional properties (Kato, 2007):

Odorants should allow leaks to be detected without any external equipment, so the end user need not worry about maintaining any measuring equipment. In this way equipment failures will not cause leaks to go undetected. Consequently, odorants allow the detection of leaks in places where it may be difficult to position detectors, and they can be used in small concentrations because of the sensitivity of the human olfactory system (Kopasz, 2007).

Furthermore, a fuel-gas odorant should be easily distinguishable from smells encountered in daily life, i.e., those smells that are experienced in daily life situations and are not perceived as a foreign or unusual odor.

By contrast, a warning smell is in general an unpleasant smell that is perceived as an odor that indicates an unusual situation, clearly distinguishable from the smells in daily life. In this way, the smell acts as a warning signal (Kato, 2007).

Odorants are chemicals that stimulate the olfactoric sense. The human olfactory system is much less sensitive than that of animals, such as dogs, which are notorious for their sense of smell. Nevertheless, a human being can still detect certain odorants in concentrations in the air in the parts per trillion (ppt) range.

There are several definitions that are relevant to quantifying the odor. The threshold odor concentration is the absolute perception threshold at which a substance can be barely identified, however faint the impression. The odor recognition threshold is the concentration at which a representative odor for a certain substance can be detected (Patnaik, 2007).

There is a difference between the minimum detectable concentration and the minimum identifiable concentration. The former is defined as the lowest concentration at which 50% of the human population is able to smell something. This concentration is sometimes known as the perceptual threshold or odor recognition threshold.

In contrast, the recognition threshold is the minimum concentration at which a certain, predefined percentage of people can identify the substance coarsely. At the detection and recognition thresholds there will still be a large number of individuals who do not detect or recognize the odor. Data indicate that, as a rule of the thumb, the recognition threshold is roughly a factor of 10 higher than the detection threshold (Kopasz, 2007).

An example of a specific method for determining the perceptual threshold is as follows. A test odorant in a dish is left in an odorless chamber for a given period of time. The air in the chamber is agitated until the concentration of the test substance becomes constant, and then it is left standing for 1 min. Panelists then enter the chamber and assess the olfactory intensity on a scale of 0 to 5. This procedure is then repeated for different concentrations of the test odorant. A perceptual threshold is then obtained by determining the concentration of the odorant that corresponds to the olfactory intensity of 2, at which the smell can easily be identified as described below. The olfactory intensity is assessed on a scale of 0 to 5 (Kato, 2007):

Certain organic sulfur compounds are used for odorization because of their inherent penetrating smell. Skunk repellents contain sulfur compounds such as trans-2-butene-1-thiol and 3-methyl-1-butanethiol. Ethylmercaptan, because of its extremely low odor threshold, is the main compound used as an odorant in natural gas and liquid propane for leak detection, although tetrahydrothiophene is also often used. Common odorization reagents are summarized in Table 15.5 and Figure 15.1 and Figure 15.2.

tert-Butylmercaptan is very common in single component odorant, with good soil penetration and a high resistance to oxidation. Its high freezing point means that it needs to be used in mixtures with other components.

Isopropylmercaptan has a strong odor and a good resistance to oxidation. It is usually blended with tert-butylmercaptan to depress the freezing point of the latter compound (Usher, 1999).

Alkyl sulfides are resistant to oxidation, but they do not have as strong an odor as the mercaptans, so they are not used as stand-alone odorants. They are usually added to lower the freezing point of mercaptans. The commonly used odorant blends fall into one of the following main categories (Usher, 1999):

Sulfur-containing compounds are widely known as odorants used for fuel gases, but they usually generate sulfur dioxide when the fuel gases are burned. Also, if the fuel gases are used in fuel cells, a desulfurizer must be installed to remove odorant components that would cause catalyst poisoning (Kato, 2007).

Attempts have been undertaken to provide sulfur-free gas odorizing compositions, including the use of alkyl acrylates, vinyl or alkyl ethers, n-valeric acid, ethyl acrylate, cyclohexene, and norbornene derivatives (Mansfeld et al., 2006). These odorants have certain disadvantages, however. For example, acrylic ester odorants are chemically unstable. The content of cyclohexene or ethylidene norbornene must be larger than those of mercaptans (Kato, 2007). On the other hand, nitrogen-containing odorants may cause the enhanced formation of nitrogen oxides, which are toxic and react with sunlight to form ozone.

Among the alcohols that are suitable as odorants, geosmin is a preferred compound. Geosmin actually means earth smell. It is a naturally occuring organic compound produced by microorganisms. The human nose is extremely sensitive to geosmin. Its structured name is 2,6-dimethylbicyclo [4.4.0]decan-1-ol. The structures of geosmin and other sulfur-free odorants are shown in Figure 15.3.

trans-2-trans-4-Decadienal also has a low perceptual threshold (Kato, 2007). Mixtures of acrylates and pyrazines have been proposed as sulfur-free odorants (Mansfeld et al., 2006). The odorizing compositions may contain antioxidants, e.g., butylhydroxyanisole, ionol, i.e., tert-butyl hydroxytoluene, hydroquinone monomethyl ether, and α-tocopherol to protect against undesired oxidation. Examples are shown in Figure 15.4.

Sulfur-free odorants are shown in Table 15.6. Commercial sulfur-free odorants (or compositions with a reduced sulfur content) include a mixture of ethyl acrylate and methyl acrylate (Gasodor S Free™) and a mixture of ethyl acrylate and tetrahydrothiophen (Spotleak Z) (Heimlich et al., 2008).

Sulfur-free odorants can be smelled readily, but the odor rather resembles garlic. Human common sense does not associate this type of smell with combustible gas, because people are accustomed to the mercaptan smell.

Some cities have changed their odorant in the natural gas pipelines to sulfur-free odorants, with reports of success (Wagner, 2005), but others have returned to the conventional sulfur-based odorants.

The behavior of Gasodor S Free™ during reforming of methane that has been odorized with just this odorant has been tested with respect to the application of methane in fuel gas systems. It has been verified that odorization with Gasodor S Free would not have a negative impact on its subsequent use in fuel gas systems (Hennings and Reimert, 2007).

Odorants are typically provided in liquid form, and are added to the gas where the distribution gas is taken from a main gas pipeline and transferred to a distribution pipeline. In such circumstances, the gas pressure may be stepped down through a regulator to a lower pressure. The odorants added to natural gas are extremely concentrated. Odorants such as tert-butylmercaptan and other blends are mildly corrosive and are also very noxious.

If a leak of odorant were to occur at an injection site, people in the surrounding area would assume that a gas leak had occurred with areas being evacuated and commerce being interrupted. However, if such mistakes become commonplace, people in the surrounding area will become desensitized to potential gas leaks and will fail to report them.

Three techniques are commonly used for odorizing incidents natural gas in a main distribution pipeline. A liquid odorant can be injected directly into the pipeline by a high-pressure injection pump, which pumps the odorant from a liquid storage tank into a small pipe that empties directly into the main gas pipeline. Because the odorant is extremely volatile, drops injected into the pipeline immediately disperse and spread throughout the gas in the pipeline. In this way, the drops of liquid odorant are dispersed in gaseous form within a few seconds (Marshall and Zeck, 2001).

The flow of gas in the pipeline is typically metered, so that liquid odorant can be injected periodically. For example, a few drops of odorant will suffice for a 30 m³ flow of natural gas. When the gas flow meter indicates that such an amount of natural gas has flowed through the pipe, another aliquot of liquid odorant is injected into the pipeline. This process is then repeated, even though the injection is performed periodically, the odorant diffusion within the gas provides adequate levels of odorant throughout the pipeline, assuming the time between injections is not too great.

Another odorization technique involves bypassing a small amount of natural gas at a slightly higher pressure than that of the main distribution pipeline through a tank containing a liquid odorant. This bypass gas absorbs relatively high concentrations of odorant while it is in the tank, and then returns to the main pipeline. The odorant, now volatilized, diffuses throughout the pipeline in much the same manner as described in the previous method (Arnold, 2000).

A third method for odorizing natural gas is to inject the odorant into the pipeline at a controlled rate. The system includes an odorant storage tank containing the odorant to be injected. A pressurized source of inert gas, such as nitrogen, maintains the tank at a desired positive pressure above that of the natural gas pipeline. An injection conduit communicates the odorant storage tank with the pipeline. A photooptic metering means, located within the injection conduit, meters the odorant to be injected into the pipeline (Zeck, 2006).

In an improved version of this method, the chemical is metered on a drop wise basis, with individual drops being counted as they pass through a measuring unit into the injection conduit and hence the pipeline. The measuring unit includes ultrasonic transmitters and receivers, which act as either proximity sensors or by measuring the transit time, to provide a measurement of the flow rate of the odorant on either a drop basis or in a steady state flow condition. Alternatively, liquid drops may land on the diaphragm of a piezoelectric sensor and thereby generate sound waves. These are then transmitted to an associated crystal, which generates a proportional electric charge resulting in a voltage difference between the two electrodes. The resulting voltage spikes can be counted and measured (Zeck, 2008).

Leaks in pipelines can be detected by means of a test fluid. The test fluid, a mixture of dimethyl sulfide in solvent, is injected into a pipeline. It will escape through any leak, and the odorant is released from the closed compartments (Quaife et al., 1991 and Quaife et al., 1993).

With the advent of hydrogen-based fuels cells, the odorization of hydrogen has become an issue (Kopasz, 2007). Here another problem emerges, since common odorants may have a negative impact on the performance of the fuel cells, since commercial odorants act as poisons for the catalysts used in hydrogen-based fuel cells, most specifically for proton exchange membrane fuel cells. Chemical compounds based on mixtures of acrylic acid and nitrogen compounds have been adopted to achieve sulfur-free odorization of the gas (Puri, 2007).

In the use of natural gas and other petroleum gases to generate hydrogen for fuel cell applications, sulfur-free natural or petroleum gases are needed, or else a desulfurization step must be incorporated in the reforming process, which adds further cost to hydrogen generation.

Fuel cells are sulfur intolerant due to sulfur poisoning of the noble metal catalysts used. If sulfur-containing odorants are used, it would be necessary to remove sulfur-containing materials, like mercaptan odorants, from the feed gas using materials like zinc oxide. However, some sulfur-containing materials, like thiophenes, cannot be removed by zinc oxide and may require a specific hydrodesulfurization process, using hydrogen gas, to remove sulfur.

A further complexity for hydrogen fuel comes from the nature of the hydrogen flame propagation. When gases burn in air, their flames propagate upward with greater ease than they propagate downward. This is primarily due to the natural upward convection of hot burnt gases. For petroleum gases, propane and methane, the upward and downward propagating lean limits of combustion are approximately the same.

However, for hydrogen, since they differ by a factor of 2.5, the amount of odorant needed for leak detection in hydrogen could be >2.5 times that needed for methane or propane. The higher quantity of the odorant needed for hydrogen odor detection further complicates the sulfur poisoning problems for hydrogen gas used in fuel cells (Puri, 2007).

One specific problem of odorization is odor-fading. The gas may be satisfactorily odorized at the source, but if it no longer has the necessary odor impact and intensity by the time it reaches the customer, escaping gas can remain undetected and result in a serious fire or explosion hazard. Basically three causes of fading may arise (Usher, 1999):

The presence of rust and air within a pipeline may act as a catalyst for the oxidation of mercaptans, resulting in compounds that do not smell at all. On the other hand, sulfide components are much more resistant to oxidation.

Dry gas is the easiest to odorize and does not cause odor-fade. Condensed liquids in the pipeline may absorb components of the odorant. Odor masking may also occur because of the odor imparted by any impurities present in the gas.

Odor-fading from odorized liquefied petroleum gas stored in carbon steel containers can occur by catalytic effects of the containers. To postpone this effect, the respective steel surfaces can be deactivated by treating them with a deactivating agent (Nevers, 1990) before exposure to the liquefied petroleum gas.

Examples of such deactivating agents are benzotriazole, tolyl triazole, mercaptobenzothiazole, benzothiazyl disulfide, or mixtures of these compounds (Nevers, 1987). It has been suggested that a mathematical model and adequate software should be developed to predict odorant fade (Altpeter, 1997).

If natural gas for storage in natural reservoirs is odorized with sulfur compounds, then a possible environmental impact can result. Some of the odorant is lost in the formation (Sasnanand, 1993). If the loss occurs in a reservoir adjacent to an aquifer, it could contaminate the water and cause environmental problems. When gas is drawn off, water is also often injected into the reservoir. A case was described in which the respective water had a strong characteristic odor (Girod et al., 1996). A stripping column has been recommended to overcome this problem.

Contaminated ground water can be decontaminated by reaction with iron (Huang and Lee, 1997). This technique was proposed to remedy ground water that was contaminated with ethylmercaptan in situ. Studies suggest chemical reactions with iron rather than an irreversible surface adsorption. Gas odorizers can be removed by extraction, similar to the usual glycol dehydration and desulfurization process (Jullian et al., 1997; Rojey et al., 1998).

Another cleaning process for the removal of tetrahydrothiophene uses an advanced oxidation technique, consisting of water treatment by UV radiation in combination with a dose of hydrogen peroxide (Panneman et al., 1997). It is possible to keep the concentrations of odorant and condensate in the effluent below 0.1 ppb.