The use of exergetic indicators in the food industry – A review

ABSTRACT

Assessment of sustainability will become more relevant for the food industry in the years to come. Analysis based on exergy, including the use of exergetic indicators and Grassmann diagrams, is a useful tool for the quantitative and qualitative assessment of the efficiency of industrial food chains. In this paper, we review the methodology of exergy analysis and the exergetic indicators that are most appropriate for use in the food industry. The challenges of applying exergy analysis in industrial food chains and the specific features of food processes are also discussed.

KEYWORDS:

• Sustainability
• food industry
• exergy analysis

List of symbols

	=	Water activity
	=	Exergy
	=	Specific chemical exergy
	=	Exergy destruction ration
	=	Exergetic factor
	=	Mass
	=	Number of moles
	=	Entropy generation number
η	=	Efficiency
	=	Pressure
	=	Thermal energy
	=	Universal gas constant
	=	Entropy
	=	Temperature
	=	Volume
	=	Mole fraction
IP	=	Improvement potential
RI	=	Relative irreversibility
SI	=	Sustainability index
CEC	=	Cumulative exergy consumption
CEL	=	Cumulative exergy losses
SED	=	Specific exergy destruction

Greek letters:

	=	Exergetic renewability
	=	Beta value (used in eco-ExA)
	=	Productivity lack
	=	Renewability performance indicator

Subscripts:

o	=	Environment of reference
i	=	Stream
j	=	Component
th	=	Thermal
pr	=	Pressure
e	=	Electrical
eco	=	Eco-exergy

Introduction

The global population will reach 9.6 billion by 2050 ( United Nations, 2013). The economic prosperity of a big part of the population is also expected to increase, leading to more affluent diet patterns that support a very demanding food system in terms of natural resources (Gerbens-Leenes et al., 2010). It is estimated that the current demand for phytomass production will at least double, but with the probability of attaining only 80% or less of the total theoretical potential yield due to competing claims in land usage, while a number of other reasons such as the increasing frequency of extreme climatic phenomena and water and phosphorus scarcity will also worsen the situation (Koning et al., 2008). As a result, there will be a strong pressure on natural resources, energy, and food. Thus, efficient and complete use of our resources will be of utmost importance.

The main challenges that the food industry will face are the need for better agricultural and post-harvest handling practices for more efficient food production that uses less energy and water, and for minimizing food wastage throughout the complete food chain (Ohlsson, 2014). The total amount of raw materials, water, and energy required along a food chain can be substantial depending on the type of food product produced (Ramirez, 2005). Generation of food waste is also considerable since about one-third of all the food produced globally is lost within the various steps of food supply chain. About 7%¹ of this loss is due to industrial processing (Fig. 1), which seems small, but this wasted food still translates into prodigal expenditure of energy, water, fertilizer, and land use, all spent in vain.

Figure 1. Left: Global food losses in relation to the total global food production. Right: Food processing losses in relation to the total global food losses. The values used for the estimations were adapted from the report by Gustavsson et al. (2011).

Many efforts have been undertaken to improve the sustainability of the food industry, and a number of positive developments can be observed. For example, Lee and Okos (2011) evaluated successfully different food processing systems that use less water and energy, while Alamilla-Beltran et al. (2011) identified emerging food processing technologies with promising applications such as electroporation, plasma processing, pulsed electric fields, and radiofrequency heating, among others. However, the practical implementation of sustainable improvements in the food industry is hindered by the vast product diversity, the specific and limited production periods, and the large distribution areas (Klemeš and Perry, 2008). Nevertheless, the need to produce food both effectively and efficiently will become even more profound because the continuation of unsustainable processing practices will contribute to the irreversible depletion of the Earth's natural resources.

Several methodologies have been proposed for assessing and improving the sustainability of various processes and products such as mass flow analysis (MFA; also known as material throughput analysis), energy analysis (EA), life cycle assessment (LCA), cradle-to-cradle design (C2C), and pinch analysis among others (Giampietro et al., 1994; Dalsgard and Munkoe, 2000; Kytzia et al., 2004; Braungart et al., 2007; Roy et al., 2009; Damour et al., 2012; Herrero et al., 2013).

Clearly, there is no shortage of sustainability assessment methods. Yet, the most challenging task for the scientific community is to agree on useful and operational criteria that can connect appropriately resource consumption with the generated services (Gong and Wall, 2001).

Currently, energy is the most common term used by the food industry to understand process and system performance or efficiency. According to Wall (1988), energy should be considered as an indestructible quantity that is conserved in every closed process. Upon transformation of one energy form to another, part of its initial quality is destroyed (irreversibly lost), leading to a lower, degraded quality (Maes and Van Passel, 2014). The concept of energy quality has been described by Van Gool (1980) as the possibility of energy exchange between a donating and an accepting stream. This possibility was defined by Cornelissen (1997) as the “maximum work potential of a material or of a form of energy in relation to its environment,” and it is known as available work or exergy, a term which was originally given by Rant (1956) after the Greek words “ϵξ” (out of) and “έργον” (work). Now, consensus among many authors from different scientific fields develops in using exergy analysis (ExA) as an objective methodology for assessing the efficiency and hence sustainability of processes and systems, because it is based on the first and the second law of thermodynamics, considering both the quantity and the quality of material and energy streams simultaneously without having to resort to subjective weighing factors (Berg, 1980; Hevert and Hevert, 1980; Dincer, 2002b; Szargut, 2005; Zvolinschi et al., 2007; Rosen et al., 2008; Sciubba, 2009; Wall, 2009; Dincer and Rosen, 2013a; Maes and Van Passel, 2014). The advantages of using ExA over other assessment methods have been discussed in detail by various authors, e.g. Gong and Wall (2001) and Dincer and Rosen (2013c). The basic principles, the general definitions, and the differences between energy and exergy have been discussed by Dincer and Cengel (2001) and Dincer (2002b).

BoroumandJazi et al. (2013) reviewed the applications of ExA in various industrial sectors in different countries. Luis (2013) focused on the chemical industry and showed that most of the ExA publications relate to the energy- and thermodynamics-related fields. According to Dincer and Rosen (2013c), ExA seems to be applied mainly by European companies, and one of the reasons could be their longer-term viewpoints on sustainability. The potential use of ExA in the food industry has been demonstrated by Apaiah et al. (2006b), and now the interest in research in this field is rapidly growing (Fig. 2(a)). However, the number of publications of the food industry in relation to the total number of publications, as shown by Luis (2013), is still small, indicating a need for identifying relevant research questions that can better couple ExA to food science and technology (Fig. 2(b)). Therefore, the aim of this review is to evaluate the usefulness of ExA as a sustainability assessment tool to summarize the most commonly used exergetic indicators and their application in the food industry to identify particular features food processes and chains have, and to identify possible future directions for further research.

Figure 2. (a) Number of published papers related to ExA applied in the food industry. The results are obtained after the comparison of 134 publications to the best of authors’ knowledge. (b) Total number of ExA publications are shown related to the chemical industry (as shown by Luis, 2013) and the food industry.

Applications of exergy analysis in the food industry

Most food process-related publications using ExA focus on drying technologies (66%), followed by food chains having wider boundaries (10%) and heating/pasteurization processes (6%) (Fig. 3(a)). The remaining studies deal with other processes such as heating, evaporating, and chilling, and some of these consider whole countries, including the agricultural sector (societal chains). The main drying technologies studied are related to general aspects of drying (17%), but innovative drying methods such as solar drying (19%) and the use of heat pump drying (16%) receive growing attention as well (Fig. 3(b)). Solar drying case studies have been reviewed by Panwar et al. (2012) and mathematical models for thin or thick layer solar drying have been discussed by Bennamoun (2012). El-Sebaii and Shalaby (2012) described the different types of solar dryers used in the drying of agricultural products. Heat pump drying systems were summarized by Colak and Hepbasli (2009), while Bruttini et al. (2001) suggested operational policies for exergetically sustainable freeze drying of pharmaceutical products. Dincer and Sahin (2004) proposed an exergetic model for the design of thermodynamically efficient moist solid drying operations, while the energetic and exergetic efficiencies as well as general sustainability aspects of dryers have been discussed by Dincer (2011). The reason that drying receives more attention is undoubted because it is one of the most energy-demanding processes due to the high latent heat of vaporization of water, and the inefficient use of energy in case of spray drying (Aghbashlo et al., 2013). Most of the publications studied (62%) focus on the practical applications of ExA, either on experimental rigs or large-scale equipment, while about one-third of these (30%) relate to modeling of food processes or chains, and the rest (8%) are the literature reviews that do not focus on the food industry but discuss relevant processing technologies (e.g. drying). Grassmann diagrams were used in about 17% of the publications to represent exergy flows, an aspect that can be important for visually communicating the results of ExA to non-expert stakeholders.

Figure 3. Publications of ExA applied to the food industry, which show (a) the main type of processes researched, and (b) the main drying technologies researched. The results are obtained after the comparison of 134 publications to the best of authors’ knowledge.

Exergy analysis

Exergy Analysis identifies parts within a system where most of the exergy is wasted and/or destroyed, and it can help to understand better the reasons that causes those inefficiencies (Tsatsaronis and Morosuk, 2012). A typical stepwise procedure to conduct a general ExA is described by Dincer and Rosen (2013a). The procedure for applying the ExA methodology in industrial food chains is proposed as follows:

1. Define the system boundaries of the food process or chain, including all crucial steps.

2. Determine an environment of reference, which should reflect local environmental conditions.

3. Conduct a mass flow analysis, an energy analysis, and an exergy analysis using only the most relevant forms of exergy to construct Grassmann diagrams.

4. Define and calculate thermodynamic indicators.

5. Interpret the results.

6. Propose and assess potential modifications/improvements.

7. Communicate the results.

Defining the system boundaries

The choice of system boundaries for the evaluation of different food processes or chains is an important step because it considerably affects the outcome of analysis as shown by Seckin et al. (2013). Stanek and Gazda (2014) argued that the system boundaries should be extended when renewable resources are included in the analysis to account for the origin (extraction) of natural resources. The use of broad system boundaries will give a more detailed overview of the analyzed chains, but it can lead to extensive calculations and excessive use of assumptions that complicate the analysis. On the other hand, the use of tighter system boundaries will not only simplify the analysis but will also omit the identification of the impact of potentially relevant “external” processes.

Defining the environment of reference

To calculate the available work of each stream, an environment of reference has to be selected. Several reference environment models have been developed, and are discussed by Dincer and Rosen (2013a), who mentioned that one of the most commonly used models is the natural-environment-subsystem model. In this model, the environmental temperature is adjusted to match the local geographical conditions of the system under study. This approach is used in many of the publications studied in this review. In cases where psychrometric processes or pressure differences are relevant, the humidity of the ambient air and the pressure of the reference environment should be considered as well. For example, Fig. 4(a) shows how the chemical exergy of 1-kg air changes with increasing its moisture fraction at constant environmental moisture content and at different environmental temperatures (K), while Fig. 4(b) shows how the total exergy of 1 kg of air changes with increasing its both moisture fraction and temperature at constant environmental moisture and environmental temperature. This shows that the selection of a particular environment of reference will influence the outcome of ExA of, for example, a drying process, since the exergetic contents of all relevant streams and the exergetic efficiencies in the analyzed system are calculated in relation to this environment.

Figure 4. (a) Chemical exergy of 1 kg of moist air as a function of its moisture content at a constant environmental moisture content (0.008 kg water/kg dry air) and at different environmental temperatures. (b) Contour plot of the total exergy of 1 kg of moist air as a function of its moisture content (kg water/kg dry air) and its temperature (K) at constant environmental moisture content (0.008 kg water/kg dry air) and at constant environmental temperature (298 K).

Defining the relevant forms of exergy

The most relevant relations for conducting ExA are shown in Table 1. It is a common practice to consider only the relevant forms of exergy involved in the process, which are classified into three main categories: the physical, the chemical, and the mixing exergy. The total exergetic content of a stream is the sum of all of these exergies. The physical exergy can be further categorized into thermal, pressure, kinetic, potential, and electrical exergy, of which the latter three exergies are fully convertible into work, meaning that these are equal in their corresponding energies. The chemical exergy, according to Dincer and Rosen (2013b), “represents the maximum work extractable from a system at the pressure and temperature of the reference environment [non-equilibrium state] as it changes to a system with the same composition, pressure, and temperature as the reference environment [equilibrium or dead state].” Therefore, the chemical exergy of the stream can be calculated by knowing the chemical composition of a mass stream, expressed in mass fractions, and the specific chemical exergy of each component, the values of which can be found in literature (Szargut, 1989; Wall, 2009). The mixing exergy is relevant when two or more different streams are mixed, causing a spontaneous loss in exergy.

Table 1. Indicative list of forms of exergy and formulas used for their calculation.

Form of exergy	Formulas
Total exergy of a stream
Physical exergy of a stream
Thermal exergy of a stream
Pressure exergy of an incompressible liquid stream
Pressure exergy of an gaseous stream
Chemical exergy of a stream
Mixing exergy

The physical exergy forms that are frequently involved in food processing are the thermal, the pressure, and the electrical exergies. However, the chemical exergies are typically much larger than the physical exergies. While most of the chemical exergies are usually preserved during food processing, any unused material side stream represents a significant loss on the total amount of exergy, which is generally larger than most losses due to inefficient use of physical exergy. Recently, Jankowiak et al. (2014) compared the extraction of isoflavones from okara (soymilk by-product) by water and ethanol and showed that, even though water leads to a lower yield of these bioactive components, it is exergetically more efficient than ethanol due to the loss of the latter (chemical exergy loss) during the distillation process and with the spent okara. Evidently, the full use of raw materials and all material streams involved in a system is more important than the efficient use of physical exergy (e.g. in heating, cooling, and phase changes).

Having all relevant stream exergies calculated, one can construct a Grassmann diagram. This diagram shows schematically the types of exergy flows considered in a process or a system. When the chemical exergy flows are excluded from the Grassmann diagram, the physical exergy streams can be shown better, which reveal those parts in the chain where most non-material losses occur, making the diagram an effective way of communicating the ExA results.

Use of exergetic indicators

Exergetic indicators, which address different aspects of thermodynamic performance, are useful to obtain a better understanding of irreversibilities and exergy losses in a food chain. A single exergetic indicator might not be sufficient to describe completely the thermodynamic performance of an industrial food chain. Various exergetic indicators have been used for the exergetic assessment of food processes and food chains as shown in Fig. 5, and a summarized (but not exhaustive) list with their definitions and applications is shown in Table 2.

Figure 5. Main exergetic indicators used in industrial food processes and food chains. The results are obtained after the comparison of 134 publications to the best of authors’ knowledge.

Table 2. Indicative but not exhaustive list with various exergetic indicators found in literature after the comparison of 134 publications to the best of authors’ knowledge.

Indicator	Formulae	References
Exergetic efficiency		Evaporation of tomato paste (Sogut et al., 2010), milk and bread processing (Ho et al., 1986), fish drying (Tiwari et al., 2009), extrusion of fish feed (Boyar et al., 2012), drying of rice (Pandey et al., 2012), carrot cubes (Nazghelichi et al., 2010), cereal grains (Amantéa et al., 2013), red pepper (Akpinar, 2004), mulberry (Akbulut and Durmuş, 2010), coroba slices (Corzo et al., 2008a, 2008b), corn (Syahrul et al., 2003; Fortes et al., 2009; Bolaji, 2011), wheat (Fortes, 2004; Fortes and Ferreira, 2004; Inaba, 2007; Assari et al., 2013), carrot (Aghbashlo et al., 2009), pistachio (Midilli and Kucuk, 2003), coffee (Hernández-Díaz et al., 2013; Fissore et al., 2014), chili (Sami et al., 2011), olive leaves (Celma and Cuadros, 2009; Erbay and Icier, 2009a, 2011), beef (Liu et al., 2008), mint leaves (Colak et al., 2008; Akpinar, 2010a; Boulemtafes-Boukadoum and Benzaoui, 2011), parsley leaves (Akpinar, 2011), jackfruit leather (Chowdhury et al., 2011), multilayer porous media (Prommas et al., 2010), solar greenhouse drying (Ozgener and Ozgener, 2009a), solar air drying (Tyagi et al., 2009), steam production (Saidur et al., 2010), co-generation in sugar production (Kamate and Gangavati, 2009), ethanol production (Palacios-Bereche et al., 2013), milling of cereals (Akinoso et al., 2013), pasta (Ozgener and Ozgener, 2006; Ozgener, 2007; Ozgener and Ozgener, 2009b), cheese powder (Erbay et al., 2014), ethanol production from banana and its residues (Velásquez-Arredondo et al., 2010)
Exergy losses		Drying of potato (Akpinar et al., 2005; Aghbashlo et al., 2008), pumpkin (Akpinar et al., 2006), apple (Akpinar, 2005b), apricot (Akpinar and Sarsilmaz, 2004), strawberry (Akpinar, 2007), pomegranate (Motevali and Minaei, 2012), wheat (Syahrul et al., 2002), multilayer porous media (Prommas et al., 2010), non-hygroscopic porous media (Prommas et al., 2012), eggplant (Akpinar, 2005a), plants (Gungor et al., 2011c)
Improvement potential		Drying of broccoli (Colak et al., 2010; Icier et al., 2010), pasta (Colak et al., 2013), cheese powder (Erbay and Koca, 2012a), plums (Hepbasli et al., 2010a, 2010b), green olives (Colak and Hepbasli, 2007), fruits and vegetables (Gungor et al., 2011a, 2011b), laurel leaves (Kuzgunkaya and Hepbasli, 2007), tinning of fish (Quijera et al., 2013)
Entropy generation		Pasteurization and concentration of orange juice (Balkan et al., 2005; Waheed et al., 2008), spray drying of milk powder (Jin and Chen, 2011), spray drying microencapsulation of fish oil (Aghbashlo et al., 2012, 2012a, 2012b, 2012c), waste heat utilization for frying potato (Aneke et al., 2012), solar drying of fruits and vegetables (Lamnatou et al., 2012)^2
Exergy destruction ratio		Drying of soybeans (Ranjbaran and Zare, 2013), plums (Erbay and Hepbasli, 2013), olive leaves (Erbay et al., 2010), cheese powder spray drying (Erbay and Koca, 2012a), sugar production (Bapat et al., 2013)
Sustainability index		Fruits and vegetables cultivation (Hepbasli, 2011), drying of laurel (Erbay and Hepbasli, 2014), cheese powder spray drying (Erbay and Koca, 2012b)
Cumulative exergy consumption		Sugar production (Raghu Ram and Banerjee, 2003; Moya et al., 2013), fish feed formulation (Draganovic et al., 2013), recycling of used cooking oil (Talens Peiró et al., 2008; Talens Peiró et al., 2010), general agricultural production (Wall, 1990; Hoang and Rao, 2010), vegetable oil production (Özilgen and Sorgüven, 2011)
Cumulative exergy losses		Sugar production (Raghu Ram and Banerjee, 2003), societal chains that include the agricultural sector (Wall, 1990; Gong and Wall, 1997).
Exergetic factor		Drying of plums (Hepbasli et al., 2010), olive leaves (Erbay et al., 2010), laurel (Erbay and Hepbasli, 2014), cheese powder spray drying (Erbay and Koca, 2012b)
Specific exergy destruction		General chilling process (Ducoulombier et al., 2007)
Relative irreversibility		Drying of olive leaves (Erbay et al., 2010), drying of laurel (Erbay and Hepbasli, 2014), cheese powder spray drying (Erbay and Koca, 2012b)
Lack of productivity		Cheese powder spray drying (Erbay and Koca, 2012b)
Exergetic renewability		Sugar production (Bapat et al., 2013)^3 (Moya et al., 2013)^4, fruits and vegetables cultivation (Hepbasli, 2011)
Renewability performance indicator		Ethanol production from banana and its residues (Velásquez-Arredondo et al., 2010), production of sugar and ethanol (Pellegrini and de Oliveira Junior, 2011)
Eco-exergy		Fish feed formulation (Draganovic et al., 2013)

¹ Also known as Cumulative degree of perfection and its use is demonstrated in the production of yoghurt (Sorgüven and Özilgen, 2012) and vegetable oil (Özilgen and Sorgüven, 2011).

² Use of the entropy generation number, N_s, which is calculated based on the ratio of the thermal energy of the product over the solar energy absorbed, and it indicates the entropy produced during the conversion of solar energy to thermal energy.

³ Defined as exergy renewability ratio:

⁴ The non-renewable fraction can also be calculated as:

• Exergetic efficiency is one of the most frequently used indicators for the sustainability assessment of food processes. It shows how well the exergetic inputs are utilized within the process or chain. It is always lower than the energetic efficiency because it represents the deviation of the current food chain from ideality. It is thus equal to the total amount of useful exergy that emerges from the system () divided by the total amount of exergy absorbed by the system ():

In other words, it shows the loss relative to the maximum theoretical work that could be achieved by the use of processing technologies in the food chain. Consequently, its maximum achievable value is fixed based on the exergetic efficiencies of the constituent steps in the chain. Therefore, the total food chain efficiency could never reach 100% even if much more efficient technologies are used.

The exergetic efficiency can be defined in various ways and the exact definition depends on what the analyst considers as an appropriate description (Cornelissen, 1997; Stougie et al., 2002). The simultaneous use of three different exergetic efficiencies has been demonstrated in an evaporating cooling process (Dincer and Rosen, 2013d) and in an orange juice concentration process (Balkan et al., 2005). According to Valero (1998), the way the exergetic efficiency is calculated depends on the way the thermodynamic costs (exergetic contents) of inputs and products are allocated. The exergetic content should be allocated proportionally to their quantities when different products of the same quality are produced. In this case, the exergetic efficiency should be expressed according to the exergy of the output products over the total exergetic inputs (Valero, 1998). If the exergy inputs are not fully exploited and only part of them leaves the system (i.e. inputs are discarded as waste and returned to become part of the environment), the exergetic efficiency should be expressed according to the exergy of the output product over the part of the input exergy that was utilized (Valero, 1998).

In general, the calculation of the efficiency should meet a set of conditions: It should be based on relevant and influential data; should be easy to calculate, should have a practical application, and should be sensitive to changes, thus enabling a range between zero and one (Stougie et al., 2002). However, the efficiency is a ratio, and therefore a relative number that does not necessarily describe its thermodynamic performance completely. For example, Fig. 6 shows that the exergetic efficiency of two different food processes is the same, however, in food process A, a considerably larger amount of exergy is lost. Therefore, the exergetic efficiency should always be explicitly defined and considered along with other thermodynamic indicators.

Figure 6. Grassmann diagrams of two different food processes that have the same exergetic efficiency.

• The second most used indicator in the publications is the absolute exergy loss. Certain exergy losses are associated with the transformation of raw materials into final products within the food chain. These losses relate to different mechanisms such as heat transfer in thermal processing, induction of phase changes, concentration, and mixing. These can be expressed directly by thermodynamic indicators (e.g. cumulative exergy losses and exergetic efficiency), and visualized by Grassmann diagrams as the decrease in the size of arrows going in and out of the system.

According to Sciubba (2009), exergy loss is a “proper indicator of the global conversion performance of an energy-conversion chain, including complex structures.” Exergy loss refers to both the exergy destroyed irreversibly within a process (internal losses) and to any other exergy that is wasted to the environment due to other inefficiencies, e.g. from waste streams or by lack of proper insulation (external losses) (Szargut et al., 1988; Valero, 1998).

Further insights on the performance of a thermodynamic process can be obtained by the advanced exergy analysis (Morosuk et al., 2013). According to this methodology, the exergy destruction of a process is split into endogenous losses (due to the operation of a component of the process in real conditions when the rest of its components run in ideal conditions) and exogenous losses (calculated by subtracting the endogenous losses from the overall exergy destruction), as well as in unavoidable losses (that cannot be improved by any technological or economic improvement) and avoidable losses (calculated by subtracting the unavoidable losses from the overall exergy destruction). Szargut (1980) proposes a dependency of exergy losses within the different parts of a multistage process, meaning that a modification in one part of a chain might reduce the local losses but could considerably influence the total losses. The importance of considering the total chain of processes instead of focusing on a single unit operation has been demonstrated experimentally in a milk processing system (Fang et al., 1995).

• The third most commonly used indicator is the improvement potential (IP). Van Gool (1997) argues that the improvement potential should be used for comparing different processes of different scales and even of different sectors, even though the obvious maximum improvement for a given process is its total exergy loss.

• The fourth most commonly used indicator is entropy generation, which is related to exergy destruction through the Gouy–Stodola relation (Duhem, 1889; Gouy, 1889a, 1889b, 1889c). Exergy destruction and entropy generation should be considered as parallel (and not opposite) concepts because the former gives information about the work that was irreversibly lost during a process in relation to a reference environment, while the latter marks the uncertainty (or disorder) in the quality of energy that is created during the utilization (degradation) of this useful work, and both are the expressions of the second law of thermodynamics (Kay, 2002).

• The fifth indicator is the exergy destruction ratio, which is also known as the depletion number, and it was originally defined by Connelly and Koshland (1997) as the exergy destroyed in a process over the total exergetic input. The exergy destruction ratio indicates a better efficiency with a lower value, contrary to most other indicators. The exergy destruction ratio is the reciprocal of the sustainability index, SI, as proposed by Rosen et al. (2008), which shows how a change in the exergetic efficiency affects the sustainability of a process.

• The cumulative exergy loss is the sixth most used indicator and is defined as the summation of the losses that occur during the production of a certain or multiple products (Szargut, 1987). The cumulative exergy losses can be calculated by subtracting the total useful exergy delivered at the last step of the chain (or throughout the chain) from the cumulative exergy consumption (Szargut, 1988).

Many other indicators have been developed in different scientific fields and have found application in the food industry but not to a large extent. The use of renewable and non-renewable energy sources can be of relevance in a thermodynamic analysis because both of these energy sources might have the same exergy content but different overall thermodynamic impact, as suggested by Stougie and Van der Kooi (2011). Recently, Maes and Van Passel (2014) introduced the renewability fraction that is useful in identifying the actual exergetic value of renewable sources, and considers the sunlight required to produce the renewable resource, the forest abatement costs for carbon dioxide sequestration and oxygen production, and the actual sunlight captured by the process studied. A similar indicator has been used in assessing the performance of strawberry cultivation in greenhouses by different heating methods (Hepbasli, 2011). Dewulf et al. (2000) introduced the exergetic renewability defined as the renewable exergy fraction used over the total exergy input, and the environmental compatibility defined as the total exergy input over the total exergy input plus the exergy required to abate emissions and wastes. Another promising indicator is eco-exergy, a concept developed by Jørgensen (2007), in which the embodied information in living organisms in the form of DNA is assigned as potential energy work. Other less used indicators are the specific exergy destruction (Ducoulombier et al., 2007; Tambunan et al., 2010; Catton et al., 2011), the exergy loss rate (Pandey and Nema, 2011), the exergy-to-energy ratio (Quijera and Labidi, 2013), the exergy heating effectiveness (Akpinar, 2010b), the weighted mean overall exergetic efficiency (used in a country scale system) (Ahamed et al., 2011; Xydis et al., 2011), the exergetic factor, productivity lack, and relative irreversibility (Xiang et al., 2004), and peak exergy (used in solar drying analysis) (Kumar et al., 2012; Cuce and Cuce, 2013).

Communicating the results of an exergy analysis

Conveying the main outcomes of ExA to non-expert stakeholders can be as important as the analysis itself. Grip et al. (2011) stated that the lack of strategy in working with ExA, the lack of information on the opportunities that it offers, the lack of competence within the organization, the lack of time, and different prioritization strategies, hindered the implementation of ExA in Swedish companies. In addition, companies perceived ExA as a method that was not required or was not applicable, or it was difficult to use and to communicate within their hierarchy levels due to asymmetric knowledge on the topic. It is clear that the communication aspect of ExA should be seriously considered during industrial sustainability assessments.

Thermodynamic sustainability

A process can be considered sustainable in thermodynamic terms when the amount of exergy lost is small during its operation, which results in most of the selected thermodynamic indicators attaining their optimal values. For example, the cumulative exergy losses and the specific exergy losses should be as low as possible, while the total exergetic efficiency should be the highest possible. In other words, the total thermodynamic price to run the process and to produce one unit of product should be minimal, while the total exergy throughput should be as efficient as possible without degrading its quality.

However, some of the indicators may show conflicting results in practice. For example, Aneke et al. (2012) compared two industrial food chillers that make use of waste heat thermodynamically: the first one being an organic rankine cycle-powered vapor compression refrigeration process, and the second one an ammonia–water absorption refrigeration system. They found that for pressure ratios higher than or equal to the breakeven point (where the coefficient of performance is identical for both processes), the first process was more efficient. However, for lower pressure ratios the second process was more efficient even though it produced higher irreversibility. The authors assigned this paradox to the fact that the absorption refrigeration process included more heat exchangers that are also entropy generators. Such conflicting results require a more in-depth observation of all the obtained values of the indicators, and the most relevant one for the studied system should be selected to assess its thermodynamic performance.

Leites et al. (2003) pointed out general rules for the design of thermodynamically efficient chemical processes, which could also be applied within the food industry. An important rule is that the use of high quality energy should be avoided in processes that demand low quality energy (Shukuya, 2013a).

An industrial system could improve its exergetic sustainability by avoiding the generation of waste material or heat streams by re-using those streams, and by making use of renewable energy sources. For example, a feasibility study showed that it is both sustainable and profitable to recover cryo-thermal exergy from a liquefied natural gas regasification process for deep freezing of agro-food products in the surrounding industries and for space conditioning in residential and commercial areas nearby (La Rocca, 2011).

Many recent publications focus on the exergetic assessment of drying processes that use solar energy or heat pumps. The advantages of using solar, wind and geothermal power were discussed by Koroneos et al. (2003), who showed that in some cases they can be more efficient than non-renewable energy sources. For example, Le Pierres et al. (2007) demonstrated that deep-freezing of foods is possible by utilizing solar, low-grade energy. Hermann (2006) quantified the global exergy resources and stressed that it is possible to meet global demands in the reduction of energy consumption by best utilizing all known exergetic reservoirs and flows available in our biosphere.

It can be concluded from the above information that food chains in the future should be designed in such a way that:

1. waste generation is avoided, minimized, or re-used, and that the complete raw materials are converted into valuable and useful products;

2. exergy destruction during processing is minimized; and

3. renewable energy sources are used instead of fossil sources.

Current challenges and future trends in designing sustainable food chains

Several important issues have to be considered when using ExA for more efficient and sustainable food production. First and foremost, the quality and safety of the final product(s) should be guaranteed in any change in a supply chain or processing step. This should be seen as a constraint on any modification that can be proposed. Second, ExA has been developed primarily in the energy conversion and chemical processing industry, and thus will need further development and should gain acceptance in the food industry. We will now shortly discuss these aspects.

• Dealing with product safety, product-process interactions, and nutritional aspects in ExA. Process optimization within the food industry is not straightforward. Even if more sustainable processing technologies can be identified by using ExA, the same should comply with safety, legislation, and consumer quality criteria. In addition, the structure of foods, both at macro- and micro-scale, is of great importance to the bioavailability of nutrients and the sensorial quality of the product. For example, the digestibility and metabolism of dietary fatty acids is affected by both their structure and state (Michalski et al., 2013). ExA does not reflect the physico-chemical transformations of different food ingredients that occur during processing (e.g. gelatinization of starch), and therefore it says nothing about their nutritive value (Dincer, 2002b).

The application of the concept of exergy on human metabolism and food consumption has been mentioned by Szargut (2005). The conversion of food in the human body releases heat equal to its lower heating value; however, only a part of this initial chemical energy content is used to run all the complex biochemical processes and most of the heat produced is lost to environment (Shukuya, 2013a). By using calorimetric data, it was shown that adenosine triphosphate (ATP) hydrolysis is the limiting factor for obtaining the maximum available nutritive exergy, and approximately 60% of this exergy is chemically bound within the human body in the form of ATP (Mady, 2013). Mady (2013) analyzed the energy conversion processes within the human body with exergy to develop health performance indicators, while a general procedure for calculating the value of human exergy consumption was given by Shukuya (2013b). The above studies show that nutrition is an important factor to consider when designing a food product. They signify the need for extending ExA to include the impact of physico-chemical transformations of food components on their exergetic nutritive quality along the total food chain.

• Dealing with industrial emissions and waste streams. The exergy value for useful streams is by definition always positive even if their main physical variable, e.g. temperature, is lower than the environment of reference (Shukuya, 2013a). However, the assignment of exergetic values to waste streams is currently a matter of debate. This issue deserves attention since ExA often deals with lost work from waste streams, and often suggests their avoidance or their reuse even when that is not possible. All streams that are brought at equilibrium with the environment of reference have zero exergy. This means that the useful work potential of a process stream that is dispersed into the environment will become part of this environment, and thus by definition has no exergy anymore. Streams that can cause harm to the environment (e.g. emissions or waste streams) have an exergetic content (e.g. thermal and/or chemical exergy) but cannot or should not be released in the environment as such. They should first be brought into a state that allows them to become at equilibrium with the environment without doing harm. This generally requires additional processing (e.g. wastewater treatment, chemical degradation, or even incineration) and therefore requires exergy to be spent. Therefore, harmful emissions bear an exergetic penalty as large as the exergetic investment needed to render them harmless to the environment.

However, there is still no clear agreement on how to treat the exergetic content of such waste streams. Maes and Van Passel (2014) argue that ExA cannot capture the immaterial aspects of emissions and waste streams (e.g. land degradation and biodiversity loss), which should be considered by additional metrics in a sustainability assessment.

Gaudreau et al. (2009) criticized ExA, posing that ExA is not objective due to the vagueness of the methodology in addressing the impact of waste streams on an infinite reference environment that should actually remain unaltered. Other authors proposed that waste streams should be considered either as constrained or unconstrained, where the former are streams of value (dictated often by economic factors), and the latter are free to influence the environment but have the potential of becoming constrained (valuable) (Dincer, 2002a; Rosen et al., 2008).

Zhu and Feng (2007) studied the allocation of cumulative exergy among the separation of multiple products from a stream by introducing a new parameter based on their minimum separation work. A similar approach could be useful for allocating the negative impact of waste streams based on their minimum abatement cost. Nevertheless, whether the exergy content of a toxic or contaminated stream that is released to the environment should be zero, or it should be allocated based on a minimum abatement cost, or even attain a negative value, is a topic that still requires attention.

• Need for a systematic framework and communication standards. The importance of the integration of ExA in industrial practice, policymaking, taxation, and education has been stressed by many authors (Van Gool, 1997; Tsatsaronis and Cziezla, 1999; Gong and Wall, 2001; Dincer, 2002b). Companies and governmental organizations are more familiar with the use of footprints. The exergy footprint was proposed by Caudill et al. (2010), which could assist in decision-making. However, these types of concepts are still not standardized and integrated to reflect the environmental, economic, and social aspects of sustainability (Čuček et al., 2012).

Hernando and Hector (2013) demonstrated the use of a framework that combines ExA with a quality control model based on Hazard Analysis of Critical Control Points (HACCP) guidelines, on the Andean blackberry cold chain. Such a systematic framework could enhance the implementation of ExA by non-expert stakeholders in the future food industries as part of exergy preservation programs that could be used as mandatory and legislative requirements of governmental sustainability projects.

• Dealing with variability in data. Oftentimes, systems with immense system boundaries are assessed by ExA, and therefore the analysis has to rely on data that are not readily available and can be found only in literature or by using expert knowledge. This implies that the analysis can convey some degree of uncertainty due to variability in literature data. Besides, the analysis outcome is strongly dependent on the model assumptions. Therefore, a consensus among the scientific community has to be reached for defining an appropriate methodology for reliable model validation, sensitivity, and uncertainty analysis.

• Method extension. ExA is continuously extended to include economic and environmental aspects. Maes and Van Passel (2014) give an overview of such methodologies such as the cumulative exergy content developed by Szargut et al. (1988), the extended exergy accounting developed by Sciubba (2001), the ecological cumulative exergy consumption developed by Hau and Bakshi (2004), and the cumulative exergy extraction from the natural environment developed by Dewulf et al. (2007). Tsatsaronis and Morosuk (2012) discuss other exergy-based methods such as exergoeconomics, exergo-environmental analysis, and the advanced exergy analysis. The latter methodology has been applied on the drying of herbs and spices by a gas engine heat pump successfully (Gungor et al., 2013). The combination of the objective power of ExA with methods stemming from other fields such as operations research can lead to the development of useful decision-making tools for cases where conflicting objectives (e.g. profit and sustainability) occur. This combination has been demonstrated in the design of a falling film evaporator (Nishitani and Kunugita, 1983), and that of a novel protein food chain (Apaiah et al., 2006a). Later on, a new graphical method, which identifies the optimum operating parameters of a distillation column and visualizes exergy losses in 3D, was introduced by Khoa et al. (2010). Further, Vintila (2012) developed an inverse analysis method that accurately measures mass, heat, and exergy transfer coefficients essential for describing transfer phenomena in transient multiphase systems.

Artificial neural networks have also been used for predicting the exergetic performance of fish oil microencapsulation by spray drying (Aghbashlo et al., 2012c). Their optimal topology for predicting the energy and exergy in a fluidized bed dryer was determined by using response surface methodology integrated with a genetic algorithm (Nazghelichi et al., 2011) in both static and recurrent modes (Nazghelichi et al., 2011). Response surface methodology was used in combination with ExA for determining the optimal process conditions for the drying of olive leaves (Erbay and Icier, 2009b) and herbal leaves (Karimi et al., 2012), and for identifying the main factors that affect the performance of thin layer drying of pomegranate arils (Nikbakht et al., 2013).

A more detailed method that extends ExA by considering the coupling of driving forces to minimize entropy production (i.e. exergy destruction) through the use of non-equilibrium thermodynamics was proposed by Kjelstrup et al. (2004). A related field is that of finite time thermodynamics which aims at elucidating the most optimal thermodynamic path or mode of operation of processes that produce the minimum amount of entropy (or destroy a minimum amount of exergy) (Andresen, 2011). The potential of those methodologies seems very exciting but their application within the food industry is yet to be explored.

Conclusions

Exergy analysis is a methodology to assess the sustainability of food chains based on objective thermodynamic laws. The results of ExA do not provide a direct solution but lead to a better understanding of the reasons for the occurrence of losses. Although ExA in the food industry is still in its infancy, it shows a growing trend with most of the applications targeting on drying processes due to the high-energy requirements involved in those processes. Exergetic indicators can be used to provide insight for potential improvements along the complete food chain. The most commonly used ones in the food industry are the exergetic efficiency, the absolute exergy loss, the improvement potential, the entropy generation, the exergy destruction ratio, the exergetic factor, and the cumulative exergy losses. A food chain is thermodynamically sustainable when the selected exergetic indicators attain their most optimal values. Each process along the food chain should be designed to utilize all the available quality of its input(s), and to degrade it in the best possible manner, i.e. destroying the least amount of exergy while generating the minimum amount of entropy by avoiding the production of waste streams, or reusing them in case where avoidance is not possible. However, when waste streams are to be re-used, the proper allocation of their exergetic content should be considered carefully as this is still a matter of debate among the scientific community. Moreover, care should be taken when defining system boundaries because these can considerably affect the outcome of analysis. Replacing fossil fuel energy sources with renewable energy sources will also contribute in improving the exergetic sustainability of a food chain.

This review identifies several points of attention for ExA to gain acceptance in the food industry. First, it is clear that any modification in the design of a food chain should comply with quality and safety standards, and any impact of physico-chemical transformations occurring during processing of food components on their nutritive quality should be quantifiable. Second, there is a need for the scientific community to reach a consensus for the appropriate use of model validation, sensitivity, and uncertainty analysis techniques whenever dealing with variability in literature data or experimental uncertainty in ExA, and therefore enhancing the robustness of the assessment. Third, the communication of the results of ExA to non-expert stakeholders can be difficult and it can be as important as the analysis itself. Considering the above points, it is clear that the acceptance of ExA by the industrial food sector as a credible sustainability assessment method will be enhanced through developing a unified framework that provides guidelines for the design of food products of maximum nutritive value by using processes that destroy a minimum amount of exergy along the complete food chain.

Highlights

• Exergy analysis can be used to assess the sustainability of industrial food chains.

• Studies that apply ExA on food processes show an increasing trend.

• Nutritive aspects will become relevant in ExA studies in the future.

• The allocation of exergetic values in industrial waste streams lacks clarity.

• A systematic framework could support ExA implementation.

Funding

The authors are grateful to the Dutch Food Retail Organization (CBL) and the Federation of the Dutch Food and Grocery Industry (FNLI) for funding this research under the project “Valorization of raw materials and process efficiency,” which is under TI Food and Nutrition, a public–private partnership on precompetitive research in food and nutrition. The public partners are responsible for the study design, data collection and analysis, decision to publish, and preparation of the manuscript. The private partners have contributed to the project through regular discussion.

Notes

¹ Calculated by using data of the report of Gustavsson et al. (2011), where the production volumes and the percentage of expected losses occurring at the processing and packaging sector for each commodity group per region were considered for the estimation of processing and packaging food losses (in million tonnes).

² Use of the entropy generation number, , which is calculated based on the ratio of the thermal energy of the product over the solar energy absorbed, and it indicates the entropy produced during the conversion of solar energy to thermal energy.