We are IntechOpen, the world s leading publisher of Open Access books Built by scientists, for scientists. International authors and editors

Transcription

1 e are IntechOen, the world s leading ublisher of Oen Access books Built by scientists, for scientists 4, , M Oen access books available International authors and editors Downloads Our authors are among the 154 Countries delivered to TOP 1% most cited scientists 12.2% Contributors from to 500 universities Selection of our books indexed in the Book Citation Index in eb of Science Core Collection (BKCI) Interested in ublishing with us? Contact book.deartment@intechoen.com Numbers dislayed above are based on latest data collected. or more information visit

2 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 1 1 X Distributed Cogeneration: Modelling of Environmental Benefits and Imact Pierluigi Mancarella and Gianfranco Chicco* Imerial College London, UK *Politecnico di Torino, Italy 1. Introduction Radical changes occurred in the energy scenario in recent years, with a clear trend towards shifting art of the energy roduction from large centralized lants to relatively small decentralized systems. The growing diffusion of distributed generation systems for Combined Heat and Power (CHP) roduction reresents a significant art of these changes. In articular, CHP generation could bring substantial imrovements in energy efficiency and energy saving, as well as economic benefits, with resect to the searate roduction () of electricity in the centralized ower system and of heat in local boilers (Horlock, 1997). The develoment of CHP systems is articularly relevant for relatively small-scale alications (e.g., below 10 M e ) in urban areas, including otential couling to heat networks for larger caacities as well as micro-cogeneration (Pehnt et al., 2006; Pehnt, 2008) for domestic alications. The adotion of CHP systems can even be more effective when it is ossible to suly, in eriods with little or no heat demand, absortion chillers to satisfy the cooling demand (for instance, for air conditioning), thus obtaining high-efficiency seasonal tri-generation systems (Meunier, 2002; Mancarella, 2006). Moreover, CHP systems can be conveniently used in a distributed multi-generation framework, to suly various tyes of chillers to better fit the overall characteristics of the demand of various energy vectors (Chicco & Mancarella, 2009a). Higher energy efficiency can also corresond to lower environmental imact in terms of CO 2 emissions with resect to, mainly deending on the generation characteristics of the ower system in the secific country (Mancarella & Chicco, 2008a), articularly where electricity generation revailingly occurs from fossil fuels. On the other hand, distributed cogeneration could worsen the air quality on the local level, due to emissions of various hazardous ollutants such as NO x, CO, SO x, Particulate Matter (PM), Unburned Hydrocarbons (UHC), and further substances conveying the ollution into the human body. In articular, in urban areas the environmental ollution is more critical because of a host of reasons, among which: a) high concentration of background ollutants, in articular due to road traffic ollution; b) difficult disersion in the atmoshere of the ollutants roduced from small-scale generators located in urban sites, with resect to large ower lants with high stacks; c) relatively high number of recetors, due to the oulation density;

3 2 Distributed Generation d) resence of relative weak recetors, such as children, elders and sick eole; and, e) detrimental effects on non-human recetors (monuments, green urban areas and ecosystems), that could also contribute to keeing the ollutants within the area. On the above reasons, the local air quality regulation could often be quite stringent, esecially in urban areas, with environmental assessments tending to be conservative and leaving reduced margins to the deloyment of cogeneration in heavily olluted zones. These limitations call for a thorough araisal at the cogeneration system lanning stage. In addition, it is imortant to consider that the emissions of certain ollutants may worsen even significantly in the off-design oeration at artial load of the cogeneration unit. Hence, the environmental assessment of cogeneration systems has to be carried out not only on the basis of the full-load erformance, but in actual oerating conditions. This asect is even more relevant considering that in today s and emerging energy systems the cogeneration units are not generally used in on/off oeration only, but can be controlled to achieve secific objectives of electrical or heating load tracking, or with more refined strategies in which the cogeneration unit is combined into trigeneration or more generally multigeneration systems (Mancarella & Chicco, 2009b) and microgrids (Hatziargyriou et al., 2007). urther environmental benefits could refer to micro-cogeneration solutions. This chater addresses the manifold sides of otential environmental benefits and imact related to modelling and analysis of distributed cogeneration solutions. Section 2 recalls the energy efficiency benefits of adoting cogeneration systems. Section 3 deals with the modelling of global and local emissions. Section 4 describes the characterization of the emissions from tyical CHP technologies already widely alied today. Section 5 resents secific indicators for environmental imact assessment. Section 6 discusses the role of environmental imact in the formulation of otimization methods. Section 7 illustrates the identification and determination of the environmental external costs from distributed cogeneration. Section 8 addresses the otential deloyment of cogeneration in energyrelated markets. Section 9 draws the conclusions and indicates directions of future research. 2. Energy efficiency of distributed cogeneration A suitable characterization of cogeneration equiment and systems is conducted by using a black-box modelling aroach, in which the erformance of CHP units is reresented by relevant inut-outut efficiency models (Mancarella, 2006; Chicco & Mancarella, 2009b). In articular, a cogeneration rime mover is reresented by using its electrical efficiency, thermal efficiency Q, and their sum known as Energy Utilisation actor (EU) (Horlock, 1997). By denoting with the electrical energy (kh e ), Q the thermal energy (kh t ), and the fuel thermal inut (kh t ) in a secified time interval, the energy efficiency indicators are exressed as: Q ; Q ; EU Q (1) The terms, Q and can also be interreted as average owers within the secified time interval. or instance, this interretation is useful for the sake of comarison of the energy roduction (reresented as average ower within a time interval in which the ower

4 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 3 variation is relatively low) with the rated ower of the equiment, to check whether the oerational limits are exceeded, rovided that significant ower variations during the secified time interval can be excluded. The fuel thermal inut is generally based on the Lower Heating Value (LHV) of the fuel. The efficiencies indicated above deend on many factors, such as the equiment technology, the loading level, the outdoor conditions, the enthaly level at which heat is roduced, the characteristics of the heat recovery system, and so forth (Danny Harvey, 2006). The outcomes of the CHP system energy efficiency assessment can be reresented in a synthetic way through suitable indicators. A classical way to define such indicators is to comare the roduction of the same energy oututs (electricity and heat) from the cogeneration system and from conventional systems for searate roduction of electricity and heat used as reference (Horlock, 1997). The systems are tyically the electricity distribution system (EDS) for electricity roduction (associated to a reference electrical efficiency e ), and conventional boilers for heat roduction (associated to a reference thermal efficiency t ). Considering the fuel thermal inut to the conventional searate roduction system, the resulting energy efficiency indicator is the Primary Energy Saving (PES), also known as uel Energy Savings Ratio (ESR), exressed as: 1 PES 1 1 (2) Q e t e Q t Energy efficiency benefits of cogeneration aear for ositive values of the PES indicator, and the break-even condition is found for PES = 0. The simle and meaningful structure of the PES indicator makes it articularly useful to quantify the energy efficiency of a cogeneration system for regulatory uroses (Cardona & Piacentino, 2005). Extensions of the PES indicator have been roosed to encomass trigeneration systems (Chicco & Mancarella, 2007a) and more general multi-generation systems (Chicco & Mancarella, 2008b). A further indicator characterising the oeration of a cogeneration system is the heatto-ower cogeneration ratio, tyically denoted with the letter lambda (Horlock, 1997), that according to (1) can be also seen as the ratio of the thermal to electrical efficiency: Q Q (3) 3. Modelling of global and local emissions 3.1 Emission factors The emissions of a generic ollutant from a combustion device can be characterised through suitable emission factors, referred to the useful energy roduced by the generic energy vector X. The corresonding emission factor model is exressed in the form m X X X (4)

5 4 Distributed Generation where the term X m reresents the mass [g] of the ollutant emitted to roduce the energy X vector X [kh], and is the emission factor or secific emissions [g/kh] of the ollutant referred to X. The emission factor deends on the tye of generator and varies in different oerating conditions (e.g., at full load or artial load), with the equiment aging and with the state of maintenance of the generator. The emission erformance of the generator can be characterised through dedicated measurements in actual oerating conditions. In the definition of the emission factors for cogeneration alications, the energy vector X can be chosen in different ways, thus originating different definitions of the emission factor. or instance, X can reresent the inut fuel energy [kh t ], or the electricity roduction [kh e ] or the heat roduction Q [kh t ]. The corresonding formulations can be written by exressing the mass of ollutant m on the basis of the emission factor adoted in different ways: m Q Q (5) In articular, the emission factor deends rimarily on the characteristics of the chemical reactions and on the tye of fuel used (Cârdu & Baica, 2002). It is then ossible to evaluate the emission factor referred to the energy roduced through the exression X (6) X where X is the efficiency equivalent to the roduction of the useful energy X using the fuel. or instance, for a cogeneration system the term X can be the electrical efficiency for electricity roduction, or the thermal efficiency Q for heat roduction. 3.2 Emission balances The emission factors can be used to formulate global or local emission balances (Mancarella & Chicco, 2009a). The global emission balance does not take into account the location of the emission source with resect to the recetors. or a given time interval (e.g., one hour) in which the variation of the energy roduction from the cogenerator is low (in such a way to assume almost constant values of the variables involved in the analysis), it is ossible to comare the mass of the ollutant emitted by the cogeneration system with the mass of ollutant emitted from searate roduction of the same electrical and thermal energy (Mancarella & Chicco, 2009a; Mancarella & Chicco, 2009b). ith reference to the electricity roduction, taking into account the cogeneration ratio (3) and considering the total mass of ollutant emitted in global searate roduction (G) as the sum of the mass of ollutant emitted in the roduction of electricity and heat, it is ossible to elaborate the revious exressions to get (Mancarella & Chicco, 2009a):

6 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 5 m G Q,, Q,, Q,, m m (7) rom (7) it is ossible to define the global equivalent emission factor referred to the electricity roduction, G Q,, (8) The emission factor, G is directly comarable with the corresonding emission factor of the cogeneration system. This is a major uside of the emission balance model develoed, which allows emission comarison on a common basis (for instance, secific emissions with resect to the same useful energy outut) and thus unbiased environmental imact araisal of different generation rationales such as cogeneration and searate roduction. In fact, by comaring the emissions er unit of useful kh roduced, the fact that cogeneration can roduce the same amount of useful energy burning less fuel (thus roducing less ollutants) is intrinsically made aarent in the model. Hence, the environmental imact benefits (in terms of reduced secific emissions) arising from enhanced cogeneration efficiency are exlicitly acknowledged as well. Conversely, comarisons carried out by considering only the concentration of ollutant (in mg/m 3 ) contained in the gases exhausted off to the ambient would be unable to take into account the energy efficiency benefits of cogeneration, as further remarked in Section 5.4. These concets form the basis to adot energy-outut related secific emissions to define the emission reduction indicators illustrated in Section 5.1. The local emission balance takes into account that the large ower lants for electricity roduction are tyically located far from urban areas. In the analysis of the effects of emitting ollutants with roagation in a relatively limited area (for instance, NO x, CO and volatile organic comounds), it is ossible to adot an aroximated model, neglecting both the roagation of these substances outside the area and the ossible introduction of the same substances emitted by sources located outside the area. or instance, for an urban area the local emission balance considers only the local searate roduction (L) from residential boilers belonging to the area. In this case, the mass of the ollutant emitted is only the one originated by the thermal roduction and is exressed as (Mancarella & Chicco, 2009a) m L Q, Q, Q, m Q (9) The local equivalent emission factor, referred to the electricity roduction, is then defined as L, L Q m, (10) More detailed evaluations can be conducted with reference to secific models of the ollutant disersion in the atmoshere (Arya, 1999). However, the local emission balance model is useful for a reliminary assessment of the local emission imact through the dedicated indicators illustrated in Section 5.2. Of course, the local emission aroximation

7 6 Distributed Generation can be more or less relevant, also deending on the secific ollutant and disersion conditions, and reresents a conservative ( essimistic ) evaluation of the actual imact due to the distributed energy system. On the other hand, the global emission aroximation may often reresent an otimistic evaluation of the imact from distributed sources (aart from the evaluation of greenhouse gases (GHGs), whose effect is essentially global to every extent, neglecting ossible contribution to micro-climates). However, the simultaneous araisal of local and global emission balances rovides meaningful insights on the uer and lower bounds of the real environmental ressure (Mancarella & Chicco, 2009a). In articular, the indications yielded by these models are indeendent of the secific site and can be used to comare different scenarios of develoment of the cogeneration systems. The information rovided can be useful for regulatory uroses, with the aim of setting u local emission limits, assuming conventional values for the searate roduction efficiencies and emission factors, as discussed in Section Characterization of the emissions from cogeneration technologies The analysis of the emissions of various ollutants from secific cogeneration technologies takes into account different tyes of fuel (mainly natural gas, or alternative fuels such as biomasses) and oeration in artial-load conditions, with otential remarkable worsening of the emission of ollutants such as NO x and CO at relatively low loading level. The illustrations included here refer to technologies that have reached a wide commercial stage, such as the ones based on microturbine (MT) or internal combustion engine (ICE) rime movers. The same concets can be extended to other technologies, such as fuel cells. Generally, different cogeneration units, also with the same tye of technology, could exhibit emission characteristics quite different from each other, because of the secific design of the combustion device, the ossible resence or abatement system, and so forth. It is then tough to draw general emission models. or general studies, it is tyically referred to consider average values of emissions taken from inventories reared by the various environmental rotection agencies and research grous worldwide (e.g., EPRI, 2009; US Environmental Protection Agency, 2009), or elaborations of data rovided by manufacturers. Oeration at artial load can be determined by the imlementation of secific control (or load tracking) strategies, such as electrical load-following or heat load-following ones, as well as economically driven strategies such as based on the evaluation of the sark sread between natural gas cost and electricity cost. The emissions in real oeration conditions deend on the characteristics of the combustion occurring in the cogeneration rime mover. Exerimental results rovided by the cogeneration unit manufacturers and obtained during secific researches and on-site measurements have shown that the emissions of some ollutants (for instance, NO x and CO) can worsen significantly during artial load oeration. urthermore, at decreasing loading level the evolution of these emissions is not linear and in some cases could exhibit nonmonotonic behaviour, esecially for microturbine units (Canova et al., 2008; Mancarella & Chicco, 2009a). These asects make the emission imact assessment of cogeneration systems more comlicated. In addition, below a certain loading level (e.g., 50%) the erformance of the cogeneration unit could become so worse to suggest switching the unit off. In this case, the domain of definition of the oerating conditions of the cogeneration unit becomes nonconnected, including the discrete switch-off condition and a continuous oeration range

8 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 7 between the technical limits of minimum and maximum loading. These asects imact on the characterization of the cogeneration systems and call for adoting dedicated analysis techniques, for instance based on mixed integer and linear/nonlinear rogramming or heuristic methods, in some cases develoed for energy efficiency analyses not including environmental asects (Horii et al., 1987; Illerhaus & Verstege, 1999; Tsay et al., 2001; Gómez-Villalva & Ramos, 2003). Concerning the tyes of analysis, one of the key distinctions occurs between time-domain simulations and methods in which the succession of the time instants in the cogeneration system oeration is not exloited. Time-domain simulations are needed when it is imortant to consider the couling-in-time of events, for instance to take into account the integral effect of the emissions within a secified eriod, or the oerational limits deendent on the time domain, such as the maximum number of switch-on/switch-off oerations of the cogeneration unit (reschi & Reetto, 2008). If the couling-in-time is not strictly relevant, general evaluations can be carried out by using integral models, such as for instance the equivalent load aroach illustrated in Mancarella & Chicco, 2009a. This aroach is based on the construction of a discrete (multi-level) version of the load duration curve reresenting the electricity demand, containing re-defined levels of artial-load oeration. Each loading level is reresented by a air hourly energy-duration (number of hours), from which the equivalent electrical load is calculated as the weighted average of the hourly energies, assuming the load level durations as weights. Each level of the duration curve is then associated to a value of secific emissions, used to determine the mass of ollutant and the equivalent emission factors for the cogeneration system and for searate roduction. The equivalent load aroach can be used for different time horizons and with different levels of detail of the load duration curve. This makes the aroach suitable both for lanning analyses over one year or more, as well as to reresent the local emissions occurring in the short-term (e.g., minute by minute) during the alication of load-tracking strategies, in order to quantify the cumulative duration for which the emission limits have been exceeded. urthermore, the structure of the equivalent load aroach rovides smooth trends of variation of the secific emissions when the equivalent load changes, even in the case of high non-linearity of the secific emissions from the cogeneration units. The equivalent load aroach could be adoted by regulatory bodies to establish a conventional technique for taking into account artial-load oeration of the cogeneration units. 5. Indicators for environmental imact assessment Secific system-based indicators are aimed at romoting olicy develoments, as well as determining the break-even conditions for which CHP systems are equivalent to the conventional searate roduction in terms of global or local emissions. The use of synthetic indicators for assessing the benefits of exloiting cogeneration technologies with resect to searate roduction is common in energy efficiency studies, as recalled in Section 2. Comarison between distributed and centralized systems can be resorted to also in terms of emission analysis (Strachan & arrell, 2006). Similar concets can be extended and alied in the framework of the global and local emission balance aroach (Section 3.2). In articular, the indicators listed in the following subsections are articularly useful to obtain

9 8 Distributed Generation general indications indeendent of the characteristic of the individual site, thus of ossible interest for regulatory uroses. 5.1 Global emission indicators Let us consider the case of CO 2 emission assessment as a relevant examle of alication of the global emission balance aroach, given the global warming imact of CO 2 as GHG. The CO 2 emission reduction due to cogeneration is exressed in relative terms with resect to the mass of ollutant emitted in global searate roduction. The resulting CO 2 Emission Reduction (ER) indicator alied to global CO 2 emissions (Mancarella & Chicco, 2009a) is a sub-case of the PER indicator introduced in Chicco & Mancarella, 2008a for multigeneration systems: m ER G CO m 2 G m, Q,, 1 1 (11) Q Q, Q where the emission balance is carried out by considering the mass m CO 2 of CO 2 emitted from the combustion of the fuel to cogenerate useful electricity and heat, and the mass G m CO 2 of CO 2 emitted by the searate roduction of the same useful oututs (electricity and heat Q) from conventional technologies. Exloiting cogeneration is environmentally effective for ositive values of ER, while ER = 0 indicates the break-even condition. As a further ste, it is ossible to introduce the CO 2 emission equivalent efficiencies (Chicco & Mancarella, 2008b),, Q,, and CO (12) 2 Q, thus exressing the ER (11) as 1 ER 1 (13), Q Q, and obtaining an exression with formal analogy with the PES indicator (2) used in energy efficiency studies. Considering the global equivalent emission factor referred to the electricity roduction defined in (8), the ER exression can be further written in terms of the electricity roduction CO 2 1, G CO 2, G CO 2 ER 1 (14)

10 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 9 Both equation (13) and the last exression in equation (14) show that the ER indicator can be exressed in terms of cogeneration efficiencies and emission factors only. The emission factor CO 2, referred to the cogeneration thermal inut, can be considered at first aroximation indeendent of the loading level, estimating it as a function of the fuel carbon content and of its LHV (Educogen, 2001). As an examle, the value CO g/kh e can be assumed for natural gas referred to the LHV. The electrical and thermal efficiencies of the cogeneration unit can be evaluated deending on the loading level, giving the ossibility of alying the indicator for exlicit assessments under actual oerating conditions of the cogeneration unit. Taking into account equation (6), a further relevant result can be obtained if the cogeneration system and the searate roduction use the same fuel. In this case, it is ossible to write equation (13) as 1 ER 1 (15) e Q t In this way, the ER indicator becomes equal to the PES indicator (2), that is, the environmental benefits can be evaluated by using only energy efficiencies, roviding emission reduction results numerically coincident with the ones obtained from the energy saving analysis, as widely discussed in Chicco & Mancarella, 2008a, and Chicco & Mancarella, 2008b. The underlying hyothesis leading to (15) is that comlete combustion occurs, which is an excellent aroximation in most cases (Educogen, 2001) and leads to a conservative model of the CO 2 emissions. In fact, with incomlete combustion art of the hydrocarbons roduce ollutants other than CO 2 (for instance, CO), so that the CO 2 roduced is lower than the one estimated by using the emission factor model. The generalisation of the indicators to assess the global emission reduction for a generic ollutant is straightforward, yielding to the class of indicators 1 1, Q,, Q, Q Q, G ER 1 (16) thus obtaining for instance indicators named NO X ER for the case of NO X, COER for the case of CO, and so forth, with the same concetual imlications described above for the CO 2 case. Concerning global warming imact, in cogeneration alications CO 2 is the main GHG of interest. However, in certain cases also methane emissions could be of concern, articularly because methane could reresent u to 90% of the total UHC emitted in natural gas-fuelled units. Thus, a further formulation is resented to enable assessing the global emission reduction for a generic GHG or for a GHG set G. This formulation is based on the fact that the effect of a generic GHG can be comared with the effect of CO 2 in terms of Global arming Potential (GP). Since by definition GP = 1 in the case of CO 2, the GP for the other GHGs is exressed in relative terms with resect to CO 2 (see also Chicco & CO 2

11 10 Distributed Generation Mancarella, 2008a, for details). The equivalent emission factor introduced for the generic GHG G is defined as X X CO eq GP 2, (17) where GP reresents the mass of CO 2 equivalent to the emission of a unity of mass of the X GHG, while is the emission factor defined in equation (6). The exression of the equivalent indicator GHGER containing the effect of a set of GHG referred to the cogeneration system is GHGER 1 CO eq 2, G, Q, CO eq, CO eq, Q 2 2 G (18) 5.2 Local emission indicators Since the local emission balance model of Section 3.2 neglects the amount of ollutants roduced by the electrical system considered to be sufficiently far from the area of interest, the local emission reduction indicators are defined starting from the global emission indicators and deleting the amount referred to the searate roduction of electricity. The class of generalised local emission reduction indicators for a generic ollutant is then exressed as 1 1 Q, Q, Q Q, L LER 1 (19) where the last exression is obtained by taking into account the definition of the local equivalent emission factor referred to the electricity roduction in (10). The corresonding indicators are named NO X LER for the case of NO X, COLER for the case of CO, and so forth. 5.3 Conventional searate roduction efficiencies and emission factors The results obtained from the alication of the emission reduction indicators deend on the choice of the efficiencies and emission factors referred to searate roduction of electricity and heat. The rationale for setting u the conventional values has to be addressed in a systematic way, highlighting the imlications of different choices that could be adoted. Different settings of the values could lead to different numerical outcomes of the indicators introduced above. Since these indicators are exressed in relative values (er unit or er cent), the relevant shareholder tyically ays attention to the resulting numerical outcome to get an idea of the otential emission reduction. or instance, a numerical outcome of 20% emission reduction has different meanings deending on the set of searate roduction efficiencies used to determine it.

12 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 11 Among the different ways to set u the conventional reference values, it is ossible to mention: 1) The definition of the conventional values on the basis of the average values of the emission factors, that is,, for electricity roduction and Q, for heat roduction. In this case, these values are assigned by considering on the electrical side the average emissions from the ower lants used to roduce electricity, and for the thermal side the average emissions from different boilers, also sulied with different fuels. The average values are calculated as weighted sums of the emissions from different units with resect to the unit sizes and tyes, and can in case refer to the marginal units oerating in the bulk ower system generation scheduling. This kind of definition allows obtaining indications on the real emission reduction that could occur in a given energy scenario, for instance in a given country (Meunier, 2002). 2) Considering cogeneration systems sulied by a given fuel (for instance, natural gas), the conventional values can be defined by taking into account the emission factors of technologies sulied by the same fuel (that is, in the case of CO 2, with the same carbon content and thus basically with the same CO 2 emissions er unit of burned fuel). This aroach is aimed at assessing the emission saving otential of CHP systems intrinsic in the lant characteristics. It is then ossible to adot the model (6) with searate roduction efficiencies e and t for electricity and heat, resectively, to determine the equivalent emissions. In turn, the searate roduction efficiencies can be chosen according to different rationales, taking into account: a) average technologies for sizes similar to the one of the cogeneration system under analysis (ASST Average Same-Size Technologies); or, b) the best available technologies for sizes similar to the one of the cogeneration system under analysis (BSST Best Same-Size Technologies); or, c) the best available technologies (BAT) without size limits (with natural gas, corresonding to high-efficiency boilers and combined cycle ower lants subtracting the electricity transmission and distribution losses). The numerical values of the searate roduction efficiencies and emission factors are given by system-wide assessments, and need to be udated after some years in order to account for ossible changes in the energy generation mix. An examle of values referred to average CO 2 emissions in the Italian system (year 2003) yields, Q, 525 g/kh e and 275 g/kh t (Mancarella & Chicco, 2009a). On the basis of equation (6), it can be noted that when increasing the efficiencies in the ER indicator cogeneration loses cometitiveness with resect to searate roduction. In fact, the adotion of the best technologies as references clearly enalizes the numerical outcome of the indicator, making cogeneration look less convenient. However, if the indicators are used for regulatory uroses, for instance setting thresholds above which it is ossible to obtain incentives, the regulatory body can set the thresholds taking into account the concetual meaning of the reference values. The above aroaches are useful to boost the investments into highefficiency cogeneration systems, with ossible economic incentives within nation-wide energy and environmental olicies (Euroean Union, 2004; Cardona & Piacentino, 2005).

13 12 Distributed Generation 5.4 Emission limits and romotion of energy efficiency The emissions measured on-site, or suitable emission reduction indicators, are tyically comared with the limits to the various ollutants established by regulatory bodies. The rationale for setting u the emission limits can lay a key role to romote or limit the diffusion of energy efficient technologies. or instance, the emission limits could be established considering the concentration of ollutant contained in the gas released to the ambient (e.g., exressed in mg/m 3 ). This aroach could intuitively seem suitable to avoid exceeding the emission thresholds. However, it is not adequate to romote energy efficiency. In fact, a generator with high efficiency and a given concentration of ollutant in the exhaust gases would be enalized with resect to another generator less efficient but with slightly lower concentration of ollutant in the exhaust gases, regardless of the fact that the actual emissions er unit of outut of the generator with higher efficiency could be lower than for the other unit. A viable alternative consists of setting u emission limits on the basis of the secific emissions X referred to the useful energy outut (for instance, exressed in mg/kh). In this case, it is ossible to romote both reduction of the real environmental imact (referred to the useful oututs) and increase of energy efficiency. Another limiting factor to the develoment of cogeneration solutions is the way in which the interactions among different causes of ollution are taken into account in the environmental regulation, esecially at the local level. In the resence of a remarkably high level of ollution due for instance to road traffic, the strict alication of the emission limits when lanning the installation of new cogeneration systems would make it hard to romote the diffusion of new efficient technologies introducing (even relatively low) new local emissions, since the burden of exceeding the emission limits would be totally charged to the marginal lants to be installed. Promoting the diffusion of energy efficient technologies at the lanning stage thus requires a comrehensive re-assessment of the causes of ollution and the identification of measures for limiting the imact of each of these causes. 5.5 Indicators for comarative emission assessment Generally, the fuel adoted to suly the cogeneration system is different with resect to the fuel considered to reresent the searate roduction, esecially when taking into account the mix of fuels used to roduce electricity in the ower lants at regional or nationwide level. Thus, focusing on CO 2 emissions, on average the same cogeneration technologies can be effective in terms of emission reduction in a system with revailing roduction of electricity from fossil fuels (above all if with heavily olluting marginal lants), while they could rovide no benefit in systems with revailing roduction from hydroelectric or nuclear sources (Meunier, 2002; Chicco & Mancarella, 2008b). These asects have been outlined in Mancarella & Chicco, 2008a, by introducing additional environmental imact indicators on the basis of which it is ossible to rovide a quantitative assessment of the effectiveness of adoting a certain tye of cogeneration within a given regional or national context. In articular, the use of the indicators denoted as CO 2 Emission Equivalent Efficiency (EEE) and CO 2 Emission Characteristic Ratio (ECR) enables the determination of the break-even conditions for which CHP systems are equivalent to the conventional searate roduction in terms of GHG emissions. The ECR indicator is defined as

14 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 13 ECR (20) The ER exression (11) can be rewritten by exlicitly showing the term ECR as ECR ECR ER 1 1 (21) Q Q Q,, Q,, It is then ossible to define the indicator EEE as the value of ECR obtained by alying the break-even condition ER = 0 (Mancarella & Chicco, 2008a): EEE Q, CO, Q,, Q 2 Q (22) In this way, the indicators ECR and EEE can be easily calculated on the basis of the emission factor of the fuel used, of the efficiencies of the cogenerator and of the emission factors in searate roduction. Adoting cogeneration is then convenient in terms of reducing the CO 2 emissions if the following inequality holds: EEE ECR (23) In analogy to equation (6), for searate roduction it is ossible to define the exressions,, e, Q, ; CO (24) 2 t Assuming that the same fuel is used to suly the cogeneration unit, the external boiler and the ower system generation mix, the EEE indicator can be exressed in a way deending only on the cogeneration unit and searate roduction efficiencies: EEE e t Q (25) In the general case, in which the fuels are not the same, the general scheme of analysis can be alied with some ractical adjustments. or this urose, suitable correction factors can be defined. Considering the emission factor,, e electricity searate roduction system, and the emission factor for the equivalent fuel sulying the,, t the boiler for searate roduction of heat, the correction factors are defined as for the fuel sulying

15 14 Distributed Generation CO 2 e ;,, e t,, t CO 2 (26) Thus, the emission factors referred to the searate roduction of electricity and heat become, resectively,,, e, CO 2 CO 2 e ee,, t, CO Q 2 ; CO (27) 2 t t t and the exression of the EEE indicator becomes EEE e t e t Q (28) thus highlighting the use of the corrected electrical efficiency e e for searate roduction of electricity, and of the corrected thermal efficiency t t for searate roduction of heat. Comarisons at regional or nation-wide level by using the ECR and EEE indicators have been reorted in Mancarella & Chicco, 2008a. 5.6 Emission mas In order to reresent effectively the outcomes from break-even analyses, secific emission maing models can be introduced. or instance, it is ossible to work out the break-even conditions (in terms of ollutant emissions) with reference to the unitary roduction of electricity or heat, in terms of emission factors as defined in equation (5), taking the electrical and thermal efficiencies of the CHP rime mover as variables. The relevant break-even emission values so obtained in these emission break-even mas can be then comared to the actual emissions from every secific cogeneration unit and for every given ollutant. In this way, it is straightforward to estimate the environmental imact of cogeneration on the basis of the relevant emission balance considered, so that the mas drawn allow for general (not only break-even) emission assessment, as illustrated below. or secific analyses it could be ossible to simulate the actual disatch of the relevant cogeneration units to assess unctually their environmental imact with resect to the marginal lants oerating in the bulk ower system, as done for instance in Hadley & Van Dyke, 2003, or for bigger systems in Voorsools & D haeseleer, 2000, and Voorsools & D haeseleer, Aart from the secific emission balance considered, several different mas can be drawn deending uon the reference emission characteristics assigned to the searate generation. The conventional reference values can be set u as indicated in Section 5.3. The reference numerical values considered for searate roduction may change significantly the outcomes of the analysis, so that these values must be carefully selected according to the secific systems under study and to the secific goal to ursue, above all for olicy uroses. Alternative models could be develoed to account for the marginal oeration of cogeneration units in different hours in an equivalent fashion, as done for instance by Tsikalakis & Hatziargyriou, However, for general and synthetic assessments, such as

16 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 15 for general olicy regulation develoment, simulation-based or equivalently detailed aroaches seem less feasible, above all in the resence of large ower systems. Once drawn the relevant break-even emission mas, given the cogeneration efficiencies (1) and the corresonding emissions for every oerating oint of the cogeneration system under analysis, it is ossible to evaluate the local and global emission balances (Section 3.2), and thus the environmental erformance with resect to the conventional searate generation of the same amount of electricity and/or heat, according to the emission evaluation model used. More secifically, given a certain ollutant, to which corresonds a certain emission break-even ma on the basis of the searate roduction emission references selected for the analysis, the environmental imact comarison between conventional generation and cogeneration can be carried out on the basis of the following stes: assign the electrical and thermal efficiency of the analysed cogeneration system under a determined oeration condition; in corresondence of these values of and Q, determine on the (local and/or global) emission break-even ma the relevant (local and/or global) emissions from the conventional searate roduction technologies taken as references; comare the emission values found to the actual emissions from the considered cogeneration system under the same oerational conditions; on the basis of the searate roduction emissions and the actual cogeneration emissions, evaluate the relevant emission balance. A secific examle of use of the emission mas is shown here, considering the case of NO x with average reference emission factors, = 0.5 g/kh e for electricity generation and NO x Q, NO x = 0.5 g/kh t for heat generation. In the local emission balance analysis, ig. 1 shows the emission break-even curves, in terms of secific emissions exressed in mg/kh e. The curves are drawn in function of the cogeneration electrical efficiency, for discrete values of the thermal efficiency used as the curve arameter. The maing in ig. 1 can be exloited by following the rocedure outlined above, with reference to secific cogeneration rime movers. or instance, let us consider a MT and an ICE with rated characteristics indicated in Table 1. unit rated ower NO x secific emissions [k e ] Q y [mg/kh e ] [mg/kh t ] MT ICE Table 1. Rated characteristics for MT and ICE units. The location within the ma of the actual secific emissions (in mg/kh e ) of the MT and the ICE units is reorted in ig. 1. or each unit, the rated electrical and thermal efficiency air of values corresonds, on the ma, to the emission break-even condition, that is, the maximum secific emissions (referred in this case to the kh e ) that the unit should feature in order to guarantee an environmental imact lower than the conventional heat generation reference in the local emission balance. In articular, while the emission ma in ig. 1 is drawn with reference to the secific NO x emissions er kh e, the break-even conditions are

17 16 Distributed Generation worked out so as to comare the cogeneration heat roduction with the same thermal energy roduced by local boilers. Indeed, an equivalent emission break-even ma might be drawn as well with reference to secific emissions er kh t roduced. As the last ste in the evaluation, once ointed out in the ma the relevant break-even unitary emissions, the local emission balance can be grahically worked out by ointing out the actual emissions from the machine under analysis. Considering the MT, the values of electrical efficiency (0.29) and thermal efficiency (0.48) are entered in the emission ma, roviding a break-even emission factor value of about 330 mg/kh e. Comaring the break-even emission value to the actual full-load MT emissions of about 170 mg/kh e (Table 1), the MT brings NO x emission reduction er kh e roduced of about =160 mg/kh e. A similar calculation could be readily develoed with resect to the kh t roduced. In this case, the MT would emit 101 mg/kh t (Table 1), to be comared with 200 mg/kh t emitted by the reference boiler; this brings NO x emission reduction equal to 97 mg/kh t, again corresonding to =160 mg/kh e. Considering the ICE, entering the electrical efficiency (0.34) and the thermal efficiency (0.49) in the emission ma, the break-even NO x emissions value for the reference boiler is of about 280 mg emitted er kh e of electrical energy cogenerated. Considering actual secific emissions of 1500 mg/kh e (Table 1), the local emissions using the ICE unit increase with resect to the local emissions from conventional reference boilers by about 1220 mg/kh e actual ICE NOx secific emissions [mg/khe] MT actual break-even break-even Q electrical efficiency ig. 1. Local NO x emission balance assessment with MT and ICE (average searate roduction references). The same considerations aly to the global emission ma for NO x, comaring the breakeven conditions with the actual emission characteristics of the MT and ICE of Table 1. In this case, a similar reresentation (not shown here) can be adoted, in which the global emission balance yields a global emission reduction of about 650 mg/kh e for the MT, and a global emission increase of about 750 mg/kh e for the ICE. Besides single cogeneration units, the emission maing models can be used to evaluate more general solutions, such as articular scenarios of diffusion of cogeneration with

18 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 17 different tyes of equiment. The general aroaches introduced enable to undertake scenario analyses aimed at assessing the environmental imact from given tyes of equiment, number of units, and so forth. or instance, the scenario analyses can be formulated according to the lines indicated in Chicco & Mancarella, 2008c: different technologies are taken into account, with their energy and emission characteristics; a set of scenarios is defined, each of which is characterized by a given mix of technologies, that can be envisioned for the future, and by the level of enetration of cogeneration with resect to the (electrical and thermal) energy demand; each scenario also contains a secific set of reference values for searate roduction of electricity and heat; for each mix of technologies, the equivalent energy and emission erformance is determined by weighting the contribution of each technology to the overall mix with a redefined model (e.g., linear); the PES and ER indicators, as well as the local emission balance outcomes, can be used to evaluate the energy efficiency and the environmental imact for different ollutants. In addition, a comrehensive environmental imact maing of given cogeneration systems can be obtained by merging scenario analyses with off-design assessments. 6. Role of environmental imact in the formulation of otimisation methods Minimization of the emissions from cogeneration lants has been included in the formulation of lanning and oeration roblems in various literature studies. The simlest way to take into account emissions is to include the emission limits within the otimization roblem constraints. More generally, emissions are taken into account in the definition of multi-objective otimisation roblems of different tyes: 1. Problems in which an equivalent objective function is defined as the weighted sum of a set of objective functions; one of the objective functions (or more than one, for instance when local and global emissions are considered searately) refers to emission imact minimization. 2. Problems with conflicting objectives (tyical roblems in the case of considering different tyes of emissions) solved through the evaluation of comromise solutions with Paretofront calculation techniques. The Pareto front contains all the non-dominated solution oints of the multi-objective otimization roblem. ith reference to objective minimisations, a solution is non-dominated if no other solution exhibits lower values for all the individual objective functions. Non-dominated solutions in which none of the individual otima is achieved may be of interest because of roviding comromise alternatives among the various objectives. or roblems with non-convex Pareto front, the -constrained method (Yokoyama et al., 1988) otimizes the referred objective by introducing the other objectives as constraints and leaving a margin of accetable solutions bounded by a user-defined threshold. Concetually, the comonents of the Pareto front could be obtained more extensively by varying the threshold. More recently, some literature methods have been roosed to find directly a number of comromise solutions belonging to the best-known Pareto front (Shukla & Deb, 2007).

19 18 Distributed Generation 3. Long-term or design roblems solved through a multi-criteria aroach (Giannantoni et al., 2005; Caraneto et al., 2007). hen the level of uncertainty becomes very large, the decision-maker can assume a wider discretion in defining a set of scenarios to be considered. These scenarios are then analysed by means of multi-attribute decisionmaking aroaches, roviding multile results, among which the decision-maker can determine the referred solution by evaluating the Pareto front through a suitable numerical technique, as described in Li, 2009, or exloiting risk-based tools (Caraneto et al., 2008). Some samle references including environmental asects in cogeneration otimization are recalled here. Curti et al., 2000, introduce in the objective function a secific term reresenting the ollution cost rate determined for each ollutant, deending on the emission level, the secific damage cost referred to the ollutant, and a user-defined enalty factor. In the multi-objective aroach with minimization of cost and multile emissions resented in Tsay, 2003, the emissions of each ollutant (CO 2, SO x, and NO x ) are modelled as a function of the fuel enthaly deendent on the emission factor. In Aki et al., 2005, otimum energy ricing is obtained as a Pareto solution for a multi-objective model considering both CO 2 emissions and economic imact on consumers. The CO 2 emission limits are also used as a constraint in the otimization model to minimize the individual cost to the consumers. Pelet et al., 2005, introduce the CO 2 emissions among the multile objectives of integrated energy systems, obtaining the best-known Pareto front through a dedicated heuristic. Boicea et al., 2009, determine the best-known Pareto front for a cluster of microturbines oerating with electrical load-following control strategy. Environmental objectives are also included in the otimization of tri-generation or multigeneration systems (for instance, Burer et al., 2003; Rong & Lahdelma, 2005; Li et al., 2006). 7. Identification and determination of the environmental external costs and role of Life Cycle Assessment External costs can be defined as the costs determined by the activities of a subject that do not aear in the economic balance of that subject. Another definition refers to external costs as arising when the social or economic activities of one grou of ersons have an imact on another grou and when that imact is not fully accounted, or comensated for, by the first grou. (Bickel & riedrich, 2005). or cogeneration lants, internal costs are tyically referred to construction and oeration of the lant. Cogeneration lant emissions imact on the environment and on the society because of the effects of the ollutants emitted on the human health or on other recetors (Gulli, 2006). Some of the costs to reduce the ollutant emissions, such as the ones for installing abatement systems, are included in the internal balance. The remaining costs referred to the imact on the environment and the society are external cost comonents. Environmental external costs are related to both local and global effects, and need to be comared to searate roduction externalities. External costs can generally be internalized, that is, included in the economic evaluation to comlete the environmental analysis. hen the net external cost balance is ositive in favour of distributed systems, internalization can be carried out through fiscal incentives, discounts on urchasing of roducts, taxes, relaxation of the air quality constraints, and so forth. or instance, adoting classical economic indicators (Biezma & San Cristóbal, 2006), ICE technologies could result more

20 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 19 convenient than microturbines on the basis of economic analysis, but MTs could exhibit emissions of CO and NO X lower than the ones of the ICEs in a local emission balance. By internalizing the external costs due to global emissions, the margins of convenience of ICEs with resect to MTs could decrease substantially. The analysis and assessment of external costs for energy system alications can be addressed as in the ExternE roject (Bickel & riedrich, 2005), based on a bottom-u aroach in which the final imacts of the energy roduction are tracked back to their initial causes by determining the chain of events denoted as imact athways. The four stages of the ExternE model address the following asects: 1. The descrition of the technology and characterisation of the related emissions. This requires the identification of the arameters referred to the emissions, like the flux of substances emitted in the environment and the concentration of ollutants inside these substances, as well as the stack height, and so forth. If the lanning analysis refers to comaring different alternatives without secifying the technical details (Canova et al., 2008), it is ossible to resort to the use of the emission factors within the emission balance aroach illustrated in Section The analysis of the territorial disersion of the emissions, with the objective of identifying the concentration of ollutants in the areas in which the recetors are located. The analysis is carried out by means of either statistical models based on the time series of environmental data measured in meteorology centres, or deterministic models, based on tracing the disersion of the ollutants according to theoretical reresentation of the henomena linked to their diffusion in the atmoshere (Arya, 1999). 3. The identification of the recetors and of the corresonding dose-resonse functions to assess the otential damages. The dose-resonse functions are defined in incremental terms, reresenting the increase of damage due to the increase of concentration of the ollutant. ith reference to the human health, the dose-resonse functions can reresent the increase in the number of subjects affected by a given athology as a consequence of the exosure to the ollutants reaching the areas in which the subject is located (Bickel & riedrich, 2005). 4. The determination of the economic value of the damages, taking into account ermanent or temorary damages. Permanent damages refer to effects leading to remature death, generally evaluated by means of the Years of Lost Life (YOLL) to obtain the Value of Life Years Lost (VOLY) indicator (Krewitt et al., 1998; Bickel & riedrich, 2005). Temorary damages refer to the quantification of illness, evaluated either by determining the cost incurred by the society to care the atients affected, denoted as Cost of Illness (COI), or by identifying the illingness to Pay (TP) of the individuals to avoid the occurrence or ersistence of the causes, also taking into account further asects of ersonal judgement like the oortunities lost because of illness (Dickie & Gerkin, 1989; Stieb et al., 2002). A more comrehensive aroach to the evaluation of the overall effects of introducing cogeneration systems in the energy context resorts to the concets of Life Cycle Assessment (LCA), as discussed and alied in dedicated studies (Dincer, 1999; González et al., 2003; Chevalier & Meunier, 2005; Bickel & riedrich, 2005). or distributed cogeneration alications, in the overall LCA environmental balance the construction of MTs or ICEs (in terms of materials, manufacturing rocess, transort and installation) imacts to a minor extent with resect to the energy generated by the unit during its oerational lifetime (Riva et al., 2006). This highlights the ractical imortance of

21 20 Distributed Generation studies addressing energy and environmental issues of cogeneration, esecially when comarisons are made among technologies using the same fuel, in which further asects concerning the construction and exloitation of the fuel transortation system have the same imact and thus can be removed from the analysis without changing the nature of the results. Conversely, LCA considerations could be needed when comaring technological alternatives adoting different fuels, esecially when the fuel roduction and transortation infrastructure has different characteristics with resect to the one used for natural gas, as in the case, for instance, of biomasses or hydrogen (Chevalier & Meunier, 2005; Pehnt, 2001). More generally, the inclusion of external costs in the objective functions of regional energy lanning studies could show more incisively the benefits of adoting cogeneration, as well as other energy efficient technologies, in the energy roduction system (Cormio et al., 2003). 8. Cogeneration deloyment in energy-related markets Cogeneration can already be exloited under secific tariff systems or within a cometitive electricity market structure. In addition, it could be ossible to trade energy-related commodities (Chicco & Mancarella, 2007b) relevant to cogeneration, such as: - GHG emission allowances, introduced within an emission trading framework aimed at limiting the GHG emissions from energy consumers; currently, each entity articiating in the emission trading mechanism is assigned a certain number of emission allowances, with the ossibility of trading the ositive or negative allowance sread on the relevant market (Boonekam, 2004); the unitary rice of the allowances is exressed in m.u./tonco 2 eq, where m.u. means monetary units; the allocation of CO 2 emissions for energy systems with multile roducts and multile inuts is illustrated and discussed in Rosen, Energy efficiency (white) certificates, corresonding to acknowledged rimary energy saving obtained from actions aimed at reducing the electricity and/or gas consumtion; the unitary rice y of the white certificates is exressed in m.u./toe. Currently, a limited number of relatively large actors can articiate in such markets, also deending on the country-secific alications. However, it can be envisaged that in the future articiation will be enlarged to smaller roducers. Profitability of otentially deloying CHP technologies within such market frameworks (rovided that the olicy structure allows it) should be evaluated through energy-environmental economic models. If the focus is secifically set on the cost of electricity roduction, an alication examle is illustrated in Mancarella & Chicco, 2009a, whose aroach defines an average roduction cost of electricity (in m.u./kh e ) based on the average fuel cost for electricity roduction. or this urose, the fuel comonent related to the electricity roduced is discounted by an equivalent amount relevant to the cogenerated heat in an incremental fashion, according to the classical incremental heat rate model (Horlock, 1997). The average roduction cost of electricity defined by this aroach can be comared to the actual electricity rices, roviding reliminary hints to assess rofitability of exloiting the cogeneration system. urthermore, multi-scenario analyses are run to calculate the sensitivities of electricity roduction cost to emission allowance rices, white certificate rices, gas rices, and conventional searate roduction references (Chicco & Mancarella, 2007b). The outcomes from such an exercise enlighten how the cometitiveness of distributed cogeneration could CO 2

22 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 21 increase substantially if adequate ricing or market framework were set u to acknowledge the ositive environmental externalities brought about by the enhanced erformance intrinsic in the combined roduction. 9. Conclusions This chater has illustrated a number of asects referred to the environmental imact of cogeneration systems, mainly focused on recent literature references and on the authors work. In articular, secific models for evaluation of global and local ollutants have been discussed, introducing relevant emission reduction indicators to quantify the otential benefits of distributed cogeneration relative to classical searate roduction of heat in boilers and electricity in centralised ower lants. Oenings aimed to internalise environmental externalities within an LCA framework or otential energy-related markets, have also been illustrated. Starting from the concets resented, there are several extensions for resent and future research at both theoretical and alication levels. On the technology side, the diffusion of new solutions of different tye (for instance, fuel cells) and/or sulied by different fuels (e.g., biomasses) can change the scenario of convenience and rofitability of adoting cogeneration in evolving energy systems. Other overall benefits could come from the interactions of various tyes of cogeneration rime movers with district heating, storage, and more generally multi-generation solutions for simultaneous roduction of different energy vectors. Broader availability and interaction of technologies can be accomanied by their more flexible exloitation under off-design conditions, introducing new uncertainty in the energy system analysis. Large uncertainty also aears at the lanning and design stages because of the need of making hyotheses and of constructing very different scenarios of evolution of the electricity and gas rices in time horizons sanning over at least one decade. urther elements of uncertainty are introduced by the resence of energy-related markets for trading white certificates or emission allowances, and by a continuously changing olicy framework, whose evolution deends on olitical decisions and on arbitrariness at the regulatory level. rom the oint of view of the decision-maker, helful resonses could come from the develoment of tools exloiting the concet of risk and formulating suitable strategies to hedge risks. 10. References Aki, H.; Oyama, T. & Tsuji, K. (2005). Analysis of energy ricing in urban energy service systems considering a multiobjective roblem of environmental and economic imact. IEEE Transactions on Power Systems, Vol. 18, No. 4, (November 2005) , ISSN Arya, S.P. (1999). Air ollution meteorology and disersion, Oxford University Press, ISBN , New York Bickel, P. & riedrich, R. (ed.) (2005). ExternE: Externalities of Energy, Methodology 2005 udate. Euroean Communities, 2005, htt:// Biezma, M.V. & San Cristóbal, J.R. (2006). Investment criteria for the selection of cogeneration lants a state of the art review. Alied Thermal Engineering, Vol. 26, No. 5-6, (Aril 2006) , ISSN

23 22 Distributed Generation Boicea, A.-V.; Chicco, G. & Mancarella, P. (2009). Otimal Oeration of a Microturbine Cluster with Partial-Load Efficiency and Emission Characterization, Proceedings of IEEE Power Tech 2009, aer 109, ISBN , Bucharest, Romania, 28 June - 2 July 2009, The IEEE, Piscataway, NJ Boonekam, P.G.M. (2004). Energy and emission monitoring for olicy use. Trend analysis with reconstructed energy balances. Energy Policy, Vol. 32, No. 8, (June 2004) , ISSN Burer, M.; Tanaka, K.; avrat, D. & Yamada, K. (2003). Multi-criteria otimization of a district cogeneration lant integrating a solid oxide fuel cell-gas turbine combined cycle, heat ums and chillers. Energy, Vol. 28, No. 6 (May 2003) , ISSN Canova, A. ; Chicco, G. ; Genon, G. & Mancarella, P. (2008). Emission Characterization and Evaluation of Natural Gas-ueled Cogeneration Microturbines and Internal Combustion Engines. Energy Conversion and Management, Vol. 49, No.10, (October 2008) , ISSN Cardona, E. & Piacentino, A. (2005). Cogeneration: a regulatory framework toward growth. Energy Policy, Vol. 33, No. 16, (November 2005) , ISSN Cârdu, M. & Baica, M. (2002). Regarding the greenhouse gas emissions of thermoower lants. Energy Conversion and Management, Vol. 43, No. 16, (November 2002) , ISSN Caraneto, E.; Chicco, G.; Mancarella, P. & Russo, A. (2007). A decision theory aroach to cogeneration lanning in the resence of uncertainties, Proceedings IEEE Power Tech 2007, aer 592, ISBN , Lausanne, Switzerland, 1-5 July 2007, The IEEE, Piscataway, NJ Caraneto, E.; Chicco, G.; Mancarella, P. & Russo, A. (2008). A risk-based model for cogeneration resource lanning, Proceedings 10 th International Conference on Probabilistic Methods Alied to Power Systems (PMAPS 2008), aer 083, ISBN , Rincón, Puerto Rico, May 25-29, 2008 Chevalier, C. & Meunier,. (2005). Environmental assessment of biogas co- or tri-generation units by life cycle analysis methodology. Alied Thermal Engineering, Vol. 25, No , (December 2005) , ISSN Chicco, G. & Mancarella, P. (2007a). Trigeneration Primary Energy Saving Evaluation for Energy Planning and Policy Develoment. Energy Policy, Vol. 35, No. 12, (December 2007) , ISSN Chicco, G. & Mancarella P. (2007b). Exloiting Small-Scale Cogeneration in Energy-Related Markets, Proceedings IEEE PowerTech 2007, aer 319, ISBN , Lausanne, Switzerland, 1-5 July 2007, The IEEE, Piscataway, NJ Chicco, G. & Mancarella, P. (2008a). Assessment of the Greenhouse Gas Emissions from Cogeneration and Trigeneration Systems. Part I: Models and Indicators. Energy, Vol. 33, No. 3 (March 2008) , ISSN Chicco, G. & Mancarella, P. (2008b). A Unified Model for Energy and Environmental Performance Assessment of Natural Gas-ueled Poly-Generation Systems. Energy Conversion and Management, Vol. 49, No. 8 (August 2008) , ISSN Chicco, G. & Mancarella, P. (2008c). Environmental Sustainability of Distributed Cogeneration Systems, Proceedings IEEE Melecon 2008, aer t5-sf0030, ISBN , Ajaccio, rance, May 5-7, 2008, The IEEE, Piscataway, NJ

24 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 23 Chicco, G. & Mancarella, P. (2009a). Distributed Multi-Generation: a Comrehensive View. Renewable and Sustainable Energy Reviews, Vol. 13, No. 3 (Aril 2009) , ISSN Chicco, G. & Mancarella, P. (2009b). Matrix Modelling of Small-Scale Trigeneration Systems and Alication to Oerational Otimization. Energy, Vol. 34, No. 3 (March 2009) , ISSN Cormio, C.; Dicorato, M.; Minoia, A. & Trovato, M. (2003). A regional energy lanning methodology including renewable energy sources and environmental constraint. Renewable and Sustainable Energy Reviews, Vol. 7, No. 2,(Aril 2003) , ISSN Curti, V.; von Sakovsky, M.R. & avrat, D. (2000). An environomic aroach for the modeling and otimization of a district heating network based on centralized and decentralized heat ums, cogeneration and/or gas furnace. Part I: Methodology. International Journal of Thermal Sciences, Vol. 39, No. 7, (July 2000) , ISSN Danny Harvey, L.D. (2006). A handbook on low energy buildings and district energy systems. Quicksilver Drive, ISBN , Sterling, USA Dickie, M. & Gerkin, S. (1989). Benefits of reduced mordidity from air ollution control A survey, Chater 6 in Valuation methods and olicy making in environmental economics, olmer H. & van Ierland E. (editors), , Elsevier Science, ISBN , Amsterdam, The Netherlands Dincer, I. (1999). Environmental imacts of energy. Energy Policy, Vol. 27, No. 14 (December 1999) , ISSN Educogen (2001). The Euroean Educational Tool on Cogeneration, December 2001, web site (available October 2009): EPRI (2009). Distributed Energy Resources Emissions Survey and Technology Characterization, Euroean Union (2004). Directive 2004/8/EC of the Euroean Parliament and of the Council of 11 ebruary 2004 on the romotion of cogeneration based on a useful heat demand in the internal energy market and amending Directive 92/42/EEC. Official Journal of the Euroean Union, ; L 52: 50 60, htt://euroa.eu.int/eurlex/ri/en/oj/dat/2004/l_052/l_ en df reschi,. & Reetto, M. (2008). Artificial Immune System in the management of comlex small scale cogeneration systems. In Handbook of research on Artificial Immune Systems and Natural Comuting: Alying Comlex Adative Technologies, Hongwei Mo (Ed.), , Medical Information Science Reference, ISBN Giannantoni, C.; Lazzaretto, A.; Macor, A.; Mirandola, A.; Stoato, A.; Tonon, S. & Ulgiati, S. (2005). Multicriteria aroach for the imrovement of energy systems design. Energy, Vol. 30, No. 10, (July 2005) , ISSN Gómez-Villalva, E. & Ramos, A. (2003). Otimal Energy Management of an Industrial Consumer in Liberalized Markets. IEEE Transactions on Power Systems, Vol. 18, No. 2, (May 2003) , ISSN González, A.; Sala, J.M.; lores, I. & Lóez, L.M. (2003). Alication of thermoeconomics to the allocation of environmental loads in the life cycle assessment of cogeneration lants. Energy, Vol. 28, No. 6, (May 2003) , ISSN

25 24 Distributed Generation Gulli,. (2006). Social choice, uncertainty about external costs and trade-off between intergenerational environmental imacts: the emblematic case of gas-based energy suly decentralization. Ecological Economics, Vol. 57, No. 2, (May 2006) , ISSN Hadley, S.. & Van Dyke, J.. (2003). Emissions benefits of distributed generation in the Texas market. Reort reared for the Gas Technology Institute. Oak Ridge National Laboratory, Oak Ridge, Tennessee, USA, htt:// Hatziargyriou, N.; Asano, H.; Iravani, R. & Marnay, C. (2007). Microgrids. An overview of ongoing research, develoment, and demonstration rojects. IEEE Power and Energy Magazine, Vol. 5, No. 4, (July/August 2007) 78-94, ISSN Horii, S.; Ito, K.; Pak, P.S. & Suzuki, Y. (1987). Otimal lanning of gas turbine co-generation lants based on mixed-integer linear rogramming. International Journal of Energy Research, Vol. 11, No. 4, (October/December 1987) , ISSN X Horlock, J.H. (1997). Cogeneration-Combined Heat and Power (CHP). Krieger, ISBN , Malabar, L Illerhaus, S.. & Verstege, J.. (1999). Otimal oeration of industrial CHP-based ower systems in liberalized energy markets. Proceedings IEEE Power Tech.1999, ISBN , aer BPT , (29 August-2 Setember 1999), The IEEE, Piscataway, NJ Kreider J.. (ed.) (2001). Handbook of Heating, Ventilation and Air Conditioning, CRC Press, ISBN , Boca Raton, L Krewitt,.; Hurley,.; Trukenmüller, A. & riedrich, R. (1998). Health Risks of Energy Systems. Risk Analysis, Vol. 18, No. 4, (August 1998) , ISSN Li, H.; Nalim, R. & Haldi, P.-A. (2006). Thermal-economic otimization of a distributed multi-generation energy system - a case study of Beijing. Alied Thermal Engineering, Vol. 26, No. 7, (May 2006) , ISSN Li, X. (2009). Study of multi-objective otimization and multi-attribute decision-making for economic and environmental ower disatch. Electric Power Systems Research, Vol. 79, No. 5, (May 2009) , ISSN Mancarella, P. (2006). rom cogeneration to trigeneration: energy lanning and evaluation in a cometitive market framework. Doctoral thesis, Politecnico di Torino, Torino, Italy, Aril Mancarella, P. & Chicco, G. (2008a). Assessment of the Greenhouse Gas Emissions from Cogeneration and Trigeneration Systems. Part II: Analysis Techniques and Alication Cases. Energy, Vol. 33, No. 3 (March 2008) , ISSN Mancarella, P. & Chicco, G. (2008b). CO 2 Emission Reduction from Multigeneration Systems: Evaluation of Combined Cooling Heat and Power Scenarios. Energetica, Vol. 56, No. 8 (August 2008) , ISSN Mancarella, P. & Chicco, G. (2009a). Global and Local Emission Imact Assessment of Distributed Cogeneration Systems with Partial-load Models. Alied Energy, Vol. 86, No. 10 (October 2009) , ISSN Mancarella, P. & Chicco, G. (2009b). Distributed Multi-Generation Systems. Energy Models and Analyses, Nova Science Publishers, ISBN , New York Meunier,. (2002). Co- and tri-generation contribution to climate change control. Alied Thermal Engineering, Vol. 22, No. 6, (Aril 2002) , ISSN

26 Distributed Cogeneration: Modelling of Environmental Beneits and Imact 25 Pehnt, M. (2001). Life-cycle assessment of fuel cell stacks. International Journal of Hydrogen Energy, Vol. 26, No. 1, (January 2001) , ISSN Pehnt, M. (2008). Environmental imacts of distributed energy systems The case of micro cogeneration. Environmental Science & Policy, Vol. 11, No. 1, (ebruary 2008) 25-37, ISSN Pehnt, M.; Cames, M.; ischer, C.; Praetorius, B.; Schneider, L.; Schmacher, K. & Voß, J.-P. (2006). Micro Cogeneration. Towards decentralized energy systems. Sringer-Verlag, ISBN , Berlin Heidelberg, Germany Pelet, X.; avrat, D. & Leyland, G. (2005). Multiobjective otimisation of integrated energy systems for remote communities considering economics and CO 2 emissions. International Journal of Thermal Sciences, Vol. 44, No. 12, (December 2005) , ISSN Rosen, M.A. (2008). Allocating carbon dioxide emissions from cogeneration systems: descritions of selected outut-based methods. Journal of Cleaner Production, Vol. 16, No. 2, (January 2008) , ISSN Riva, A.; D Angelosante, S. & Trebeschi, A. (2006). Natural gas and the environmental results of life cycle assessment. Energy, Vol. 31, No. 1, (January 2006) , ISSN Rong, A. & Lahdelma, R. (2005). An efficient linear rogramming model and otimization algorithm for trigeneration. Alied Energy, Vol. 82, No. 1, (Setember 2005) 40-63, ISSN Rong, A. & Lahdelma, R. (2007). emissions trading lanning in combined heat and ower roduction via multi-eriod stochastic otimization. Euroean Journal of Oerational Research, Vol. 176, No. 3, (ebruary 2007) , ISSN Shukla, P.K. & Deb, K. (2007). On finding multile Pareto-otimal solutions using classical and evolutionary generating methods. Euroean Journal of Oerational Research, Vol. 181, No. 3, (Setember 2007) , ISSN Stieb, D.M.; De Civita, P.; Johnson,.R.; Manary, M.P.; Anis, A.H.; Beveridge, R.C. & Judek, S. (2002). Economic evaluation of the benefits of reducing acute cardioresiratory morbidity associated with air ollution. Environmental Health: A Global Access Science Source, Vol. 1, No. 1, (18 December 2002) Strachan, N. & arrell, A. (2006). Emission from distributed vs. centralized generation: The imortance of system erformance. Energy Policy, Vol. 34, No. 17, (November 2006) , ISSN Tsay, M.-T.; Lin,.-M. & Lee, J.-L. (2001). Alication of evolutionary rogramming for economic disatch of cogeneration systems under emission constraints. Electrical Power and Energy Systems, Vol. 23, No. 8, (November 2001) , ISSN Tsay, M.-T. (2003). Alying the multi-objective aroach for oeration strategy of cogeneration systems under environmental constraints. Electrical Power and Energy Systems, Vol. 25, No. 3, (March 2003) , ISSN Tsikalakis, A.G. & Hatziargyriou, N.D. (2007). Environmental benefits of distributed generation with and without emissions trading, Energy Policy, Vol. 35, No. 6, (June 2007) , ISSN US Environmental Protection Agency (2009). Catalogue of CHP Technologies,

27 26 Distributed Generation Voorsools, K.R. & D haeseleer,.d. (2000). The influence of the instantaneous fuel mix for electricity generation on the corresonding emissions. Energy, Vol. 25, No. 11, (November 2000) , ISSN Voorsools, K.R. & D haeseleer,.d. (2003). The imact of the imlementation of cogeneration in a given energetic context. IEEE Transactions on Energy Conversion, Vol. 18, No. 1, (March 2003) , ISSN Yokoyama, R.; Bae, S.H.; Morita, T. & Sasaki, H. (1988). Multiobjective generation disatch based on robability security criteria. IEEE Transactions on Power Systems, Vol. 3, No. 1, (ebruary 1988) , ISSN Acronyms ASST Average Same-Size Technologies BAT Best Available Technologies BSST Best Same-Size Technologies CHP Combined Heat and Power ER CO 2 Emission Reduction COI Cost of Illness DG Distributed Generation ESR uel Energy Savings Ratio GHG GreenHouse Gases G Global Searate Production ICE Internal Combustion Engine LCA Life Cycle Assessment LHV Lower Heating Value L Local Searate Production MT Microturbine m.u. monetary units NG Natural Gas PER Poly-generation CO 2 Emission Reduction PES Primary Energy Saving PM Particulate Matter Searate Production toe tonne of oil equivalent UHC Unburned Hydro-Carbons VOLY Value of Life Years Lost TP illingness to Pay YOLL Years of Lost Life

28 Distributed Generation Edited by D N Gaonkar ISBN Hard cover, 406 ages Publisher InTech Published online 01, ebruary, 2010 Published in rint edition ebruary, 2010 In the recent years the electrical ower utilities have undergone raid restructuring rocess worldwide. Indeed, with deregulation, advancement in technologies and concern about the environmental imacts, cometition is articularly fostered in the generation side, thus allowing increased interconnection of generating units to the utility networks. These generating sources are called distributed generators (DG) and defined as the lant which is directly connected to distribution network and is not centrally lanned and disatched. These are also called embedded or disersed generation units. The rating of the DG systems can vary between few k to as high as 100 M. Various new tyes of distributed generator systems, such as microturbines and fuel cells in addition to the more traditional solar and wind ower are creating significant new oortunities for the integration of diverse DG systems to the utility. Interconnection of these generators will offer a number of benefits such as imroved reliability, ower quality, efficiency, alleviation of system constraints along with the environmental benefits. Unlike centralized ower lants, the DG units are directly connected to the distribution system; most often at the customer end. The existing distribution networks are designed and oerated in radial configuration with unidirectional ower flow from centralized generating station to customers. The increase in interconnection of DG to utility networks can lead to reverse ower flow violating fundamental assumtion in their design. This creates comlexity in oeration and control of existing distribution networks and offers many technical challenges for successful introduction of DG systems. Some of the technical issues are islanding of DG, voltage regulation, rotection and stability of the network. Some of the solutions to these roblems include designing standard interface control for individual DG systems by taking care of their diverse characteristics, finding new ways to/or install and control these DG systems and finding new design for distribution system. DG has much otential to imrove distribution system erformance. The use of DG strongly contributes to a clean, reliable and cost effective energy for future. This book deals with several asects of the DG systems such as benefits, issues, technology interconnected oeration, erformance studies, lanning and design. Several authors have contributed to this book aiming to benefit students, researchers, academics, olicy makers and rofessionals. e are indebted to all the eole who either directly or indirectly contributed towards the ublication of this book. How to reference In order to correctly reference this scholarly work, feel free to coy and aste the following: Pierluigi Mancarella and Gianfranco Chicco (2010). Distributed Cogeneration: Modelling of Environmental Benefits and Imact, Distributed Generation, D N Gaonkar (Ed.), ISBN: , InTech, Available from: htt:///books/distributed-generation/distributed-cogeneration-modelling-ofenvironmental-benefits-and-imact