Environmental Impact of Information and Communication Equipment for Future Smart Grids

2.1.2.1 Energy analysis

It is a matter of common knowledge that it is possible to store energy within systems (e.g., in batteries). Moreover, it is possible to convert energy from one form to another (e.g., coal energy to electrical energy) and to transfer it from one system to another. In the course of all the storages, conversions, and transfers, the entire quantity of energy must be conserved [18]. This fact is embodied in the first law of thermodynamics, which states that the change of the internal energy (U) of a system is equal to the sum of heat (Q) supplied to the system and work (W) done by the same system on its surroundings [6], i.e.

Therefore, the energy balance of an entire process or system is equal to zero [19]:

According to energy balance, input and output energies of a process or system are equal. A portion of the energy at the output may be converted, e.g., into waste heat. This waste heat may as well have a positive side effect, since it could be exploited for, e.g., heating purposes. Nevertheless, the total amount of energy is conserved in all the storages, transfers, and conversions. Using an energy balance, it is possible to assess the energy demanded by a system. However, it does not tell us how well the energy is being exploited by the same system. With an energy analysis, it is not possible to determine the true thermodynamic inefficiencies related to an energy transformation system [18]. An energy analysis can identify merely the inefficiencies arising due to energy transfers out of a system that cannot be exploited anymore in the considered or some other system.

This leads to the conclusion that utilizing energy as an indicator for the assessment of energy system advantages and imperfections can be quite unclear and inaccurate [10]. An energy analysis enables quantifying energy flows. For that reason, a complete assessment of a system is facilitated. Nevertheless, even a complete life cycle energy analysis exhibits considerable disadvantages [7]. The main reason for this is the fact that an energy analysis does not consider the second law of thermodynamics. Because of that, it is not possible to directly compare distinct forms of energy with each other. However, an exergy-based life cycle assessment (E-LCA) provides means to estimate the exergy consumption over the entire considered lifetime of a system and takes the second law of thermodynamics under consideration. Based on that, it is further possible to provide information on the quality of energy, i.e., how efficiently the energy is being utilized by a system. E-LCA will be given its own discussion, and its benefits will be explained in more detail.

2.1.2.2 Life cycle assessment (LCA)

A life cycle assessment (LCA) provides means to estimate environmental effects of a product, process, or system over its respective lifetime, i.e., under consideration of its entire life cycle [20]. It was specified by the International Organization for Standardization (ISO) 14040 family of standards [14, 15]. LCA cannot be directly or precisely considered as an indicator. It rather defines a framework for indicators that can be used to evaluate how various products, processes, or systems impact the environment during their entire lifetime [13]. Based on that, various production approaches and techniques may be analyzed with regard to their environmental effects, with the aim to indicate the most efficient (i.e., environmentally sustainable) one [19]. The evaluation of environmental effects can be accomplished by capturing all inputs and outputs of a product, process, or system during its considered lifetime. Even though there are many variants of a LCA, most of them are based on emissions [13]. Figure 3 depicts schematically the life cycle assessment (LCA) framework, which is accomplished in four phases [15]:

1. Goal and scope definition

2. Inventory analysis

3. Impact assessment

4. Interpretation

The goal and scope definition phase considers the aims of the assessment and the constraints that have to be taken into consideration. The goal definition includes further the purpose of the assessment and defines the group of people potentially interested in the LCA outcomes. The scope definition defines additionally, e.g., the functional unit, data specifications, assumptions, and constraints of the assessment [15, 19]. The inventory analysis considers all inputs and outputs related to the various processes of the considered system. Inputs and outputs of the processes can be divided into economic and environmental flows [19]. Environmental inputs and outputs, which either come from, or are emitted (i.e., released) to, the environment, are considered as environmental flows. The result of the inventory analysis can be found in the inventory table. This table comprises all the resources and toxic substances that are captured or released during the considered lifetime, i.e., during the entire life cycle. Impact assessment represents the third phase of the LCA framework. During this phase the data in the inventory table is evaluated. For this purpose, impact categories are deployed, e.g., climate change, human toxicity, ozone layer depletion, and ecotoxicity [19]. In the final, i.e., interpretation, phase of a LCA, the obtained outcomes are evaluated and clarified. Based on this assessment, it is possible to suggest further improvement potentials.

A LCA provides a thorough assessment of environmental effects but has also a few drawbacks because of such a complete analysis. The first, and probably the main, disadvantage of a LCA is the lack of a simple (i.e., single) outcome which could serve as the basis for assessment and evaluation purposes between various approaches. For that reason, a direct comparison between different impact categories shows to be not that easy (e.g., ecotoxicity vs. global warming potential) [7]. Even though collections of standardized impact factors are available, the estimation of diverse environmental effects, caused from various processes, is still based on subjective assumptions and conclusions [7]. The inaccuracies introduced in the course of such vague assumptions may not be tolerable and would most probably lead to uncertain or even useless results. The second disadvantage of a LCA relates to its time exposure and accomplishment expenses. LCA software tools are existent and in use (e.g., SimaPro, GaBi Software, EarthSmart) [13, 16, 21, 22]. Still, other environmental sustainability indicators are easier to generate compared to a LCA. Exergy-based life cycle assessment (E-LCA) is an example of such an environmental sustainability indicator and will be discussed in the following part of this subsection.

2.1.2.3 Exergy-based life cycle assessment (E-LCA)

Exergy is defined as “the maximum theoretical useful work, which can be obtained from a system when brought into thermodynamic equilibrium with its environment while the system interacts with this environment only” [18]. The environment required for computing the exergy values is known as the exergy reference environment or thermodynamic environment. It is free from any irreversibility. Moreover, the exergy value of this environment must be zero. The natural environment does not meet the needs of the exergy reference environment, since it is not in equilibrium. For that reason, a model for the thermodynamic environment is always presumed. In most cases, the actual local environment is chosen as the exergy reference environment. In other words, exergy can be understood as the quantity of energy that can be utilized to perform useful work, i.e., the amount of energy available to be consumed [6]. The theory behind exergy, its characteristics, and benefits are defined by the first and second law of thermodynamics. The first law of thermodynamics defines the concept of energy conservation (see Energy Analysis). The second law of thermodynamics introduces additionally the concept of non-conservation of entropy [10, 19], i.e., the concept of entropy increase. For that reason, an exergy analysis may be utilized to evaluate, construct, and upgrade various systems. Before further discussing exergy, a few words will be addressed to entropy.

In conjunction with thermodynamic processes, entropy may be understood as the amount of energy deficiency present to perform useful work. As characterized by Rudolf Clausius, entropy can be considered as a state function of a reversible cyclic process, known as the Carnot cycle, which states that the change of entropy S is proportional to the heat Q divided by the absolute temperature T, i.e.

It is important to underline that the process assumed in Clausius definition is a reversible one. However, irreversible processes usually lead to an increase of entropy. Assuming an isolated system is subjected to some process, entropy can only be greater or equal to zero:

This relation is satisfied with equality only for a reversible process, whereas for an irreversible process, the entropy will be greater than zero. One further aspect with regard to entropy should be mentioned. It is possible to distinguish thermodynamic entropy from logical entropy [6]. According to The American Heritage Dictionary, thermodynamic entropy (expressed in Joule per Kelvin, i.e., J/K) is formulated as “the quantitative measure of the amount of thermal energy per unit temperature not available to do work in a closed system.” Logical entropy, on the other hand, can be understood as “a measure for randomness in a closed system.” The entropy concept can be applied in many fields of research, e.g., thermodynamics, communications, and statistics [6]. Even though an entropy balance provides information on inefficiencies associated with a considered system based on entropy creation, it fails to communicate the precise amount of energy being exploited by the system. The entropy concept does not provide any information about the quality of energy, just like an energy analysis. However, an exergy analysis eliminates this drawback related to an energy analysis and the entropy concept. This is because exergy allows evaluating the quality of an energy carrier (i.e., its ability to perform useful work) [18].

Using an exergy analysis instead of an energy analysis, a more thorough assessment of a system can be accomplished [20]. The data obtained from an exergy analysis contains more valuable and relevant information than the one gained from an energy analysis. Further, an increase of efficiency and means to decrease thermodynamic losses may as well be realized and determined using an exergy analysis. Moreover, an exergy analysis allows different forms of energy (i.e., with different qualities) to be compared directly with each other, as the main criterion is the capability to perform useful work. For that reason, an evaluation of various systems can be accomplished in a more accurate and meaningful way. Environmental effects and improvement potentials may as well be analyzed and assessed with the use of an exergy analysis [10]. In contrast to energy, exergy is destructed and never conserved for a majority of real processes because of irreversibilities. As opposed to an energy analysis, where the energy balance for an entire process equals zero (see the paragraph Energy analysis), the exergy balance corresponds to irreversibilities related to the process under consideration [19], so that

According to the energy balance, the difference between input and output energy is zero, i.e., they are equal (see Eq. (2)). The exergy balance, on the other hand, defines a decrease of the quality of energy (i.e., exergy) during a process [10]. This fact is obvious from Eq. (5), according to which the exergy at the input of a system is greater than the exergy at its output. The exergy balance, which corresponds to irreversibilities associated with the considered process, is proportional to the creation of entropy ∆S weighted by the temperature of the exergy reference environment T0 [6, 19]:

This relation is also known as the law of exergy loss, or the law of Gouy-Stodola [19]. The exergy loss of a complete system can be obtained by assessing and summing up the exergy losses of its corresponding subsystems or components, i.e.

Exergy losses can be divided into internal and external exergy losses. External exergy losses are composed of waste and exergies emitted (i.e., released) from a system. They encompass the amount of exergy that cannot be exploited to perform useful work. Internal exergy losses are losses related to internal inefficiencies (i.e., irreversibilities) of processes, which lead to a decrease of energy quality. They can further be classified into technical and structural exergy losses [19]. Technical exergy losses originate from system inefficiencies, while structural exergy losses are defined by various system assumptions, its composition, and features. Technical exergy losses can be minimized by applying improvement procedures. A decrease of structural exergy losses may only be achieved by modification and upgrading measures of the considered system [23].

The exergy-based life cycle assessment (E-LCA) takes all exergy inputs to a system or process during its entire lifetime into account, i.e., it includes all exergy inputs over the entire system or process life cycle [24]. Moreover, an accumulation (i.e., conservation) of exergy during the considered life cycle exergy assessment is excluded [25]. For that reason, it can be concluded that the life cycle system or process exergy losses have to be proportional to the overall exergy consumption [19, 25]. This total exergy consumption can be further used to assess and interpret the environmental sustainability of different approaches, systems, and processes.

In the following few lines, the composition of the total exergy of a system is presented. If electrical, magnetic, nuclear, and surface tension effects can be excluded, then it is possible to divide the total exergy of a system (Exsystem) into four components, namely, physical exergy (Exphysical), chemical exergy (Exchemical), kinetic exergy (Exkinetic), and potential exergy (Expotential) [18], i.e.

The physical exergy can be calculated according to Frangopoulos [18]:

Here, U, V, and S are internal energy, volume, and entropy of the system, in that order. The subscript 0 in Eq. (9) indicates the state of the considered system at temperature T0 and pressure p0 related to the exergy reference environment. The physical exergy of a system can be further divided into thermal exergy (Exthermal) related to the system’s temperature change from the temperature of the exergy reference environment and mechanical exergy (Exmechanical) introduced because of a difference of the system’s pressure from the pressure of the exergy reference environment [18], i.e.

In analogy to this, it is also possible to split the chemical exergy into reactive exergy Exreactive and nonreactive exergy Exnonreactive [18], i.e.

The chemical exergy can be understood as the hypothetical maximum of useful work that can be attained from a system as this chemically equilibrates with its exergy reference environment. In order to determine the chemical exergy, it is not sufficient to define just the temperature T0 and pressure p0 but also the chemical consistency of the exergy reference environment. Finally, the kinetic and potential exergies can be calculated according to the following Eqs. (10):

As can be seen from Eqs. (12) and (13), kinetic and potential exergies are equal to kinetic and potential energies. The variables v and h correspond to the velocity and height relative to that of the exergy reference environment (v0 = 0, h0 = 0) [18].

The preceding two parts of this subsection elaborated in detail two important but different approaches for an environmental sustainability analysis of ICTs, namely, the life cycle assessment (LCA) and the exergy-based life cycle assessment (E-LCA). In the following part of this subsection, a comparison between them is provided with the aim to point out the advantages of E-LCA for the assessment and evaluation of environmental effects, i.e., the environmental sustainability, associated with various ICT equipment.

2.1.2.4 LCA vs. E-LCA

A life cycle assessment (LCA) enables a thorough assessment of environmental effects related to ICT equipment, by capturing all inputs and outputs during its considered lifetime [26, 27, 28]. Most LCA approaches base their analysis of environmental effects on emissions, which pose a potential to negatively impact the environment, e.g., greenhouse gas (GHG) emissions. However, such an analysis brings also considerable drawbacks with it, like time exposure, accomplishment expenses, and, more importantly, the absence of a simple and unambiguous outcome for meaningful comparison purposes between different approaches. The exergy-based life cycle assessment (E-LCA), on the other hand, represents a sound approach for the environmental sustainability analysis of ICT equipment, based on its lifetime exergy consumption, which serves as a measure for the attained environmental sustainability. The major benefit of E-LCA is the fact that it leads to a single outcome (i.e., an exergy consumption value) that can be easily compared with various other potential approaches. Moreover, the quite moderate time exposure, accomplishment expenses, as well as the update and expandability features of E-LCA make this thermodynamically based indicator the best choice for the environmental sustainability analysis of ICTs.

The conclusions drawn from the E-LCA do not differ much from those of a LCA. Actually, they coincide quite well in all life cycle stages of the considered ICT equipment, the only difference being the relative scales and the quantity characterizing the environmental effects (e.g., GHG emissions vs. exergy consumption). Figure 4 shows the results of a LCA of three different smartphones, namely, an Apple iPhone 4S, a Nokia Lumia 920, and a Huawei U8652 [29]. The contribution of the different smartphone life cycle stages to the climate change, attributed to GHG emissions, over a lifetime of 3 years, is expressed in kilograms of carbon dioxide equivalents (kg CO₂e) [29, 30]. The production stage accounts for raw material extraction and processing, manufacturing and assembly, as well as the transportation during these two processing phases; the use phase is based on a 3-year operation of the smartphones; the transportation stage includes the transport of the products to their distribution location, while the recycling stage includes the transport of products to recycling plants, their separation, and shredding. It is evident that the use of a LCA for the assessment of environmental effects, even for the three considered smartphones, leads to relatively different outcomes (ranging between 16 and 70 kg CO₂e), which is a result of the quite different assumptions made for each of these approaches and LCA tools used for their assessment. Even this simple example illustrates the underlying ambiguity of a LCA. However, the conclusions drawn from each of these approaches suggest that the most dominant life cycle stage of a smartphone relates to its production stage, i.e., its raw material extraction and processing, as well as its manufacturing and assembly (including transport).

The same conclusions on the environmental sustainability are provided by means of the E-LCA. Figure 5 depicts the LCA and three E-LCA use cases (UCs) of an Apple iPhone 5C, respectively [31]. The assumptions provided in Ref. [31], i.e., those concerning the production, transportation, customer use, as well as recycling, have been adapted to a good degree into the E-LCA framework, in order to obtain a more meaningful comparison between these two environmental sustainability analysis approaches. The three depicted E-LCA UCs assume different usage intensities and therefore different daily smartphone charging durations. The first UC assumes a daily charging duration of 4 h, the second UC of 8 h, and the third UC of 12 h. An operational duration of 3 years is assumed. It should be noted that the environmental impact of the different smartphone life cycle stages in Figure 5 is given in percentage of the lifetime GHG emissions in the case of the LCA and in percentage of the lifetime exergy consumption in the case of the E-LCA for easier comparison purposes. Figure 5 suggests that for both the LCA and the E-LCA approach, the most dominant (i.e., environmentally relevant) life cycle stage of the smartphone relates to its production. The contribution of the use phase to environmental effects is more pronounced in the case of the LCA than in the case of the E-LCA. This is mainly related to the power grid mix assumed in Ref. [31] (which is accounted at a continent level), as well as the deployed LCA tool. Such an approach underlines again the ambiguity of a LCA, as different power grid mix assumptions and LCA tools will unavoidably lead to different relative outcomes (compare also the use phases of the three different smartphones depicted in Figure 4). E-LCA, on the other hand, bases its assessment of the use phase on the operational exergy consumption, leading to a more scientifically reasonable and justified approach. The difference between the use phases of the LCA and E-LCA becomes smaller assuming a longer daily charging duration (i.e., usage intensity).

For that reason, it can be argued that the conclusions drawn from a LCA and the E-LCA coincide to a good degree with each other. The main benefit of E-LCA, however, is the fact that it leads to a simple and unambiguous outcome, useful for easy and meaningful comparison purposes between various potential approaches. Such a feature lacks in a LCA, as each environmental sustainability assessment approach includes different assumptions, as well as LCA tools, leading to relatively different results (see Figure 4). Therefore, E-LCA is chosen as the environmental sustainability indicator of choice for the assessment of ICT equipment in the scope of this study.

To summarize the discussion regarding the thermodynamically based environmental sustainability indicators that can be applied to ICTs, Table 1 provides a comparison between the discussed indicator types. It can be concluded that the environmental sustainability of ICTs may at best be assessed and evaluated using the E-LCA approach. E-LCA proved to be the best choice among the other discussed thermodynamically based environmental sustainability indicators, namely energy analysis and LCA. It allows assessing the lifetime exergy consumption of various ICT approaches, which will, on the other hand, serve as an indicator for the attained environmental sustainability. Even though E-LCA does not provide detailed information of environmental effects like a LCA, it derives at a simple outcome (i.e., a single value) that can be used to compare different approaches, systems, and processes. If more information of a particular system is needed, and a more thorough assessment deems appropriate, a LCA can be applied [7]. Moreover, E-LCA satisfies the requirements defined by SCOPE for useful and applicable indicators.

E-LCA has the major benefit of being “flexible,” as it can be modified and changed when, e.g., manufacturing techniques are changed. The new data can be quite easily incorporated into the existing assessment framework. Exergy analysis can be used to assess and analyze heterogeneous systems. Since recently, E-LCA has also been used to analyze and evaluate the environmental sustainability of ICTs [7, 8, 13, 25, 32, 33].

As the final topic in this section, the relation between environmental impact, exergy efficiency, and environmental sustainability will be discussed. There are a few ways to define exergy efficiency. One of them, known as the simple efficiency, is given by the exergy at the output divided by the exergy at the input of a system or process [19], i.e.

There are also other definitions for the exergy efficiency. However, the simple exergy efficiency defined by Eq. (14) serves quite well for elaboration and assessment purposes in the scope of this study. It can be argued that an increase of exergy efficiency, i.e., a decrease of exergy losses (see also Eq. (6)), is an important step toward the improvement of the environmental sustainability of a system, process, or approach. Moreover, such an improvement would most probably lead to a reduction of the environmental impact associated with the considered system, process, or approach. Figure 6 illustrates the relation between environmental impact, exergy efficiency, and environmental sustainability. An increase of the exergy efficiency is accompanied by a decrease of the environmental impact and at the same time by an increase of the environmental sustainability. Therefore, measures to increase the exergy efficiency, or equivalently decrease the exergy losses (i.e., exergy consumption), are desirable to achieve more environmentally sustainable systems or processes. This will, moreover, result in a decrease of environmental effects.

Even though E-LCA does not provide detailed information on environmental effects in comparison to a LCA, it can be argued that a low-exergy consumption (i.e., low-exergy losses or equivalently a high-exergy efficiency) leads to less environmental effects and, more importantly, to an increase of the environmental sustainability of the considered system, product, or approach. This observation will serve as the basis for the assessment and evaluation of the attained environmental sustainability of ICT equipment deployed in the various smart grid domains. It has to be mentioned that the relation between environmental impact, exergy efficiency, and environmental sustainability depicted in Figure 6 does not hold for all processes. For instance, if a process deploys specific pollution control methods, it is possible to observe a decrease of environmental effects, even in the course of a reduction of the exergy efficiency, i.e., an increase of exergy losses or equivalently an increase of the exergy consumption.

2.2.1.1 Raw material extraction and processing

Before going into the examination of the various exergy consumption values, it will be repeated what is understood under exergy consumption. Exergy can be understood as the amount of energy available to perform useful work. For example, high-concentration ores possess an exergy value in comparison to the Earth’s crust [13]. The extraction of these ores for manufacturing and assembly (i.e., production) purposes is related to an exergy loss from the environment, i.e., exergy consumption. The estimation of the raw material extraction and processing exergy consumption of ICT equipment is achieved by using mass-specific exergy consumption values for different materials obtained from Refs. [8, 13]. For that reason, the mass of various materials, which make up the different components and devices, needs to be provided. The mass-specific exergy consumption values for raw material extraction and processing of different materials are given in Table 2.

The energy required for mining, transporting, and refining defines, among others, these raw material extraction and processing exergy consumption values. The exergy required for the extraction of materials or ores with high concentrations, compared to that of the Earth’s crust, is also included in this process stage [13]. For a lot of materials, characterized by a minor weight, mass-specific exergy consumption values were not available. For that reason, an order of magnitude estimate approach deemed appropriate and was applied to those materials in order not to magnify their share to the overall exergy consumption in this particular life cycle stage [6]. Because of the low weight of many of these materials, the deviation from the true or exact raw material extraction and processing exergy consumption value is assumed to be very low, i.e., negligible [8].

2.2.1.2 Manufacturing and assembly

The manufacturing and assembly exergy consumption is composed of the energy required by the machinery and procedures for manufacturing and assembly purposes and the exergy contained in the resulting material waste streams. It is estimated that the waste stream for metals and plastics corresponds to 10 and 50%, respectively [13]. Mass-specific exergy consumption values for the manufacturing and assembly of metals and plastics are taken from Refs. [8, 13]. The manufacturing and assembly of printed circuit boards (PCBs), integrated circuits (ICs), and processors involves very complex and more profound energy-related techniques and procedures than those applied to metals and plastics. The exergy consumption values for the manufacturing and assembly procedures of these components are as well taken from Refs. [8, 13]. The exergy consumption expended on the manufacturing and assembly of PCBs is determined and provided on a per area basis, which is based on an average dimension assumption of PCB. The manufacturing and assembly exergy consumption of ICs is determined and provided on a per IC basis, which is based on an average dimension assumption of ICs. The manufacturing and assembly of processors relies upon highly purified silicon wafers and is accompanied by large amounts of water and chemicals. This leads further to various side product waste streams and explains, moreover, the high quantity of this exergy consumption value. The overall manufacturing and assembly exergy consumption of a system, component, or device is composed of the exergy consumption portions of the various processes involved in its production. The exergy consumption values for the different manufacturing and assembly processes can be found in Table 3.

As can be seen from Table 3, the most complex manufacturing and assembly procedures (i.e., processes with the highest exergy consumption expenditure) are those for PCBs, ICs, and processors. For that reason, the estimation of the manufacturing and assembly exergy consumption of metals and plastics can be neglected for devices with a low mass, e.g., smartphones, tablets, power line communication (PLC) modems, and data concentrators, since the portion of the more complex processes is expected to dominate.

As these devices are, furthermore, considered to be equipped with quite a few ICs and processors (i.e., higher-exergy consumption-related components), the deviation from the true or exact exergy consumption in this particular process stage is expected to be very low.

2.2.1.3 Operation

The operational exergy consumption is composed of the operational power consumption, i.e., the electricity required to power the various ICT equipment, and the cooling exergy consumption, i.e., the electricity demanded by the cooling infrastructure, if present. The framework for the evaluation of the operational power consumption and the cooling exergy consumption will be discussed in more detail in the following two segments.

1. Power consumption of the networking equipment

The operational power consumption takes only the electric power (i.e., electricity) demanded by the system (i.e., some ICT equipment) into consideration, including the cooling requirements within the considered system, i.e., internal fans. The operational power consumption of a system, expressed in joule (J), can be calculated according to the following relation (modified from Hannemann et al. [8]):

where Psystem,peak, given in watts (W), denotes the system’s peak electricity consumption. Ĺsystem represents the average load of the system during its use. It is expressed in % of the peak system load. tup is the system’s daily operation time and is expressed in % of the time it is deployed. toperational denotes the system’s total usage time, expressed in years. Finally, C represents a constant required for a correct unit conversion. It is important to mention that the operational power consumption, described by Eq. (15), takes the entire exergy that enters the considered system during its use into account. It does not depict its actual destruction. Even though the heat emitted from the system may be exploited to perform useful work (e.g., for heating purposes), most systems, however, discard this emitted heat. For that reason, Eq. (15) defines the total exergy loss (i.e., exergy consumption), which is related to the electricity consumption of the considered system.

1. Power consumption of the equipment for cooling

The cooling exergy consumption may represent a major part of the operational exergy consumption. It is important for the analysis of data centers, which house a large number of servers and other equipment (e.g., switches, routers). Its power consumption is defined by the electricity demanded by, e.g., the computer room air conditioning (CRAC) units, in general cooling equipment required for a data center’s proper operation. According to Hannemann et al. [8], the cooling peak power consumption can be considered to be approximately proportional to the server’s peak power consumption. The cooling exergy consumption, expressed in joules (J), can be calculated using the following relation (modified from Hannemann et al. [8]):

PCRAC,peak, given in Watt (W), denotes the CRAC units’ peak power consumption. δdynamic denotes a binary indicator, which is used to indicate if a linear adjustment of the CRAC units’ electricity consumption to that of the servers is in force, i.e., active. If it is, it is equal to one (i.e., δdynamic = 1); otherwise it is zero (i.e., δdynamic = 0). ĹCRAC represents the average CRAC load. It is expressed in % of the peak CRAC load. This factor is needed just in the case of a dynamic cooling system (i.e., for δdynamic = 1), where an adjustment of the cooling electricity consumption to that of the servers is in force. tup corresponds to the CRAC units’ daily operation time and is expressed in % of the time it is deployed. toperational denotes the CRAC units’ total usage time, expressed in years. C also denotes here a constant needed for a correct unit conversion [6]. Equation (16) considers also here, equivalently to Eq. (15), the entire exergy that enters the considered cooling system (e.g., CRAC units) during its use, not its actual destruction. Even though the exergy contained in the CRAC unit waste emissions may be exploited to perform useful work, this amount is considered to be negligible and will not be considered in further assessments [6].

2.2.1.4 Recycling and disposal

It is not an easy task to obtain the exact amount of exergy consumed during the recycling and disposal (i.e., dismantling, shredding, separating, and recovering useful materials) of a particular ICT device. The main reason for this is the fact that accurate and trustworthy information on the goods (e.g., notebooks, smartphones) delivered to the recycling plants is not available. In Hannemann et al. [8], an order of magnitude estimate approach was assumed, according to which the amount of exergy expended on the recycling of a server (relative to its mass) corresponds to approximately 520 kilo joules per kilogram (kJ/kg). This value is assumed to be reasonable and will, therefore, be used in this correspondence as the reference for the evaluation of the recycling and disposal exergy consumption of ICT equipment.

2.2.1.5 Transportation

In order to make the exergy-based life cycle assessment (E-LCA) complete, the exergy consumed during the transportation between the various life cycle stages needs to be included in the analysis as well. After raw material extraction and processing, the materials have to be transported to the manufacturing and assembly location. From there, the final products will be transported to the location where they will be operated (or used). The final stage is the transportation of used up, damaged, or outdated components and devices to recycling plants. Three different transportation stages in the life cycle of ICT equipment will be considered, and these are material transportation, product transportation, and end-of-life transportation to recycling plants. The transportation exergy consumption does not only depend merely on the mass of the materials but also on the distance between the different process stages and the transportation mode [13]. Table 4 shows the mass (and distance)-specific exergy consumption values for the different transportation modes.

One further important fact, which concerns the provided exergy consumption values, has to be pointed out. A lot of manufactured and assembled (i.e., produced) ICT components and devices, for which diverse raw materials (e.g., aluminum, steel, plastic, copper; see also Table 2) need to be extracted and processed, will not be used for smart grid applications only. Therefore, the total estimated raw material extraction and processing exergy consumption are weighted by a usage factor (UF) of the considered ICT component or device. This is done with the aim not to overestimate the impact of the considered ICT component or device to the total exergy consumption in this particular life cycle stage. For the same reason, the estimated exergy consumption in the other process stages (i.e., life cycle stages) needs to be weighted by such a UF as well, including manufacturing and assembly, operation, recycling and disposal, as well as transportation. User devices (UDs) represent such devices, since smartphones, tablets, and notebooks will be used for other applications as well. In fact, they will be used for other purposes most of the time (e.g., various Internet services, telephony, mobile and video games). They will be used for home energy management purposes, e.g., in average only 2 h daily. To account for this fact, the consumption of smartphones, tablets, and notebooks will be multiplied by their respective UF, referred to as home energy management usage factor (HEMUF). The UF of home energy management systems (HEMSs), on the other hand, is accounted for through its daily uptime. Other examples of ICT equipment in the smart grid, which will be used for only a few minutes daily, are the Global System for Mobile Communications (GSM) and Universal Mobile Telecommunications System (UMTS) radio access network (RAN) components, i.e., base transceiver station (BTS), base station controller (BSC), Node B, and radio network controller (RNC) racks, as well as cables connecting these racks. These components will be multiplied by their respective UF as well, termed smart grid application usage factor (SGAUF), accounting for the time they are used for the advanced metering infrastructure (AMI) smart grid application (AMI will be discussed in Section 3.1.1 in more detail). The same UF will be used for core network (CN) components, i.e., serving General Packet Radio Service (GPRS) support node (SGSN) and gateway GPRS support node (GGSN) racks, core switches, as well as cables connecting this equipment. Utility equipment (UE), i.e., smart meters, power line communication (PLC) modems, and data concentrators, will be used for the AMI smart grid application only, so their respective UF is one. The same holds for the data and control center (DCC) equipment (i.e., UF = 1), which manages and controls the entire electric power grid.

3.2.1.1 Scenario 1a: influence of the utility equipment (UE) lifetime

As the customer and distribution domains of the smart grid will be equipped with a huge number of utility equipment (UE), namely, smart meters, power line communication (PLC) modems, and data concentrators, the means to gain insight into the exergy consumption of this equipment in connection with different lifetime assumptions would prove beneficial. For that purpose, three different use cases (UCs) are defined which assume different lifetimes of the considered UE. The assumptions for these three UCs are listed in Table 6.

UC 1 assumes a short lifetime of UE, i.e., smart meters and PLC modems are replaced every 5 years, and the data concentrator even every 3 years. UC 2 and UC 3, on the other hand, assume a longer lifetime of UE. UC 2 defines, moreover, the basis for Scenario 1.b, which analyzes how the number of smart meters that can be served by a single data concentrator influences the cumulative embodied and operational exergy consumption. It should be noted that the assumed number of smart meters connected to a data concentrator for the present Scenario 1.a equals to 150, which corresponds to the number provided by UC 1 of the following Scenario 1.b (see also UC 1 in Table 7). Information on the amount of data traffic per data concentrator, required for the assessment of AMI, was obtained from Luan et al. [44].

3.2.1.2 Scenario 1b: influence of the data concentrator (DC) configuration

The number of smart meters that can be served by a data concentrator can be very high. Up to 2000 smart meters can be linked to a single data concentrator [45]. The present scenario analyzes how the number of smart meters connected to a data concentrator relates to the distribution and development of the cumulative embodied exergy consumption (EEC) and operational exergy consumption (OEC) of the overall system. Two use cases (UCs) with different assumed numbers of smart meters linked to a data concentrator are considered. Table 7 provides the assumptions for these two UCs. This scenario is based on UC 2 of Scenario 1.a, according to which the lifetime of smart meters, PLC modems, and data concentrators equals to 15, 10, and 7 years, respectively (see also UC 2 in Table 6).

UC 1 assumes that 150 smart meters are linked to a data concentrator, whereas UC 2, with 2000 smart meters per data concentrator, represents the upper limit. This scenario will give more insight into the cumulative EEC and OEC in the case of these two UCs. Moreover, UC 1 of the present Scenario 1.b (with 150 smart meters per data concentrator) defines the basis for Scenario 2 described in the following subsection.

3.2.2.1 Scenario 2.a: influence of the devices’ lifetime

The present scenario can be seen as the enhancement of Scenario 1.a, as it considers additionally to utility equipment (UE) also user devices (UDs), i.e., smartphones, tablets, notebooks, home energy management systems (HEMSs), and digital subscriber line (DSL) modems, essential for a proper and easy utilization of home area networks (HANs). As in the case of Scenario 1.a, three use cases (UCs) are defined, which assume different lifetimes of the considered UDs. Each UC of Scenario 2.a is, moreover, related to the respective UC of Scenario 1.a. This can be seen in Table 8, which provides the assumed lifetimes for both the UDs and UE for the three considered UCs.

The first UC assumes a short lifetime of UDs as well as UE. UC 2 defines the most probable case considering the lifetimes of UDs and UE. UC 3 assumes an extended lifetime for both the UDs and UE, i.e., their replacement period is longer than that of the first two UCs. Based on these UCs, more information on the exergy consumption distribution and its development will be provided. UC 2 defines, moreover, the basis for Scenario 2.b as well as Scenario 2.c, which assess how different UDs’ parameters, e.g., daily charging durations and home energy management usage factors (HEMUFs) of smartphones, tablets, and notebooks, daily uptimes and average loads of HEMSs, as well as HEMS configurations, influence the cumulative embodied and operational exergy consumption. It should be noted that the assumed daily charging durations and HEMUFs of smartphones, tablets, and notebooks, daily uptimes and average loads of HEMSs, as well as HEMS configurations for the present Scenario 2.a correspond to those of UC 2 of Scenario 2.b as well as Scenario 2.c discussed in the following two parts of this subsection.

3.2.2.2 Scenario 2.b: influence of the home area network (HAN) utilization

This scenario analyzes how different daily charging durations and home energy management usage factors (HEMUFs) of smartphones, tablets, and notebooks, as well as daily uptimes and average loads of home energy management systems (HEMSs), influence the exergy consumption of the overall system. For that purpose, three use cases (UCs) are defined which assume different utilization intensities of home area networks (HANs). The assumed daily charging durations and HEMUFs of smartphones, tablets, and notebooks, as well as the daily uptimes and average loads of HEMSs for these three UCs, are provided in Tables 9–11. This scenario is based on UC 2 of Scenario 2.a, according to which the lifetime of smartphones tablets, notebooks, HEMSs, and DSL modems equals to 2, 2, 3, 4, and 5 years, respectively, and that of smart meters, PLC modems, and data concentrators to 15, 10, and 7 years.

UC 1 corresponds to a high utilization of HANs, which is associated with longer smartphone, tablet, and notebook daily charging durations and higher HEMUFs, as well as a higher average HEMS load. UC 3, on the other hand, is associated with a low utilization of HANs. UC 2 represents the most probable usage pattern of HANs. Moreover, UC 2 of the present scenario defines the basis for Scenario 2.c described in the following part of this subsection. The daily uptime of HEMSs and DSL modems is set to 100% (i.e., 24 h) for all three UCs, however, with varying average loads (see Table 11). Further, the average load of the DSL modem is assumed to be proportional to that of the HEMS for all the three considered UCs. It is important to mention that the assumed number of households that can be served by a single HEMS for the present Scenario 2.b equals to 20 households, which corresponds to the number provided by UC 2 of the following Scenario 2.c (see also UC 2 in Table 12).

3.2.2.3 Scenario 2.c: influence of the configuration of the home energy management system (HEMS)

The last scenario analyzes how the number of households that can be served by a single home energy management system (HEMS) influences the cumulative embodied exergy consumption (EEC) and operational exergy consumption (OEC) of the overall system. As already mentioned several times across this chapter, the HEMS is assumed to be implemented in the form of a 2-unit (2 U) rack-mounted server, which is placed at a convenient location near the various homes and/or buildings it is responsible for to manage. Moreover, the HEMS is considered to be equipped with all the necessary software programs required for a correct and reliable functioning of home energy management applications, i.e., home area networks (HANs). Three different use cases (UCs) are considered which assume different HEMS configurations, i.e., numbers of households it can serve. The assumptions for this three UCs are provided in Table 12. This scenario is based on UC 2 of Scenario 2.a as well as Scenario 2.b, which assume a medium lifetime of user devices (UDs) and utility equipment (UE), as well as a medium utilization intensity of HANs.

UC 1 and UC 2 assume that the HEMS can support home energy management applications of 10 and 20 households, respectively. UC 3, on the other hand, assumes that 100 households can be served by a single HEMS. Based on this scenario, more information on the cumulative EEC and OEC for the three defined UCs will be provided. This provides, moreover, means to indicate the most exergy consumption-related category in the case of these three UCs, i.e., whether the EEC or OEC dominates over the considered operational duration of 20 years.