Sustainability of electricity supply technology portfolio

Annals of Nuclear Energy 36 (2009) 409–416

Contents lists available at ScienceDirect

Annals of Nuclear Energy journal homepage: www.elsevier.com/locate/anucene

Sustainability of electricity supply technology portfolio Stefan Roth a, Stefan Hirschberg b,*, Christian Bauer b, Peter Burgherr b, Roberto Dones b,1, Thomas Heck b, Warren Schenler b a b

Axpo Holding AG, Baden, Switzerland Paul Scherrer Institut, Villigen, Switzerland

a r t i c l e

i n f o

Article history: Received 9 October 2008 Received in revised form 13 November 2008 Accepted 13 November 2008 Available online 30 January 2009

a b s t r a c t This paper outlines the approach to the evaluation of sustainability of current and future electricity supply options of interest for a major Swiss utility Axpo Holding AG. The motivation behind this effort has been to provide a solid basis for a state-of-the-art interdisciplinary assessment and use this framework within a dialog with a wide spectrum of stakeholders. The development and implementation of the methodology was coordinated by Axpo in co-operation with the Paul Scherrer Institut (PSI) and other scientific institutions. The evaluation covers environmental, social and economic dimensions of sustainability. Methods used include among others life cycle assessment (LCA), impact pathway approach (IPA) and probabilistic safety assessment (PSA). The associated databases developed by PSI have been extensively used, subject to major extensions necessary for analyzing the future technologies. Learning curves were employed for future cost estimates. Furthermore, particularly in the social area expert surveys were used. The results were aggregated using total (internal plus external) costs approach and multi-criteria decision analysis (MCDA). For MCDA a set of criteria and the associated indicators was established. In total 75 indicators were quantified, including 11 environmental, 33 social and 31 economic. Eighteen current and 18 future technologies have been analysed including nuclear as well as fossil and renewable technologies. Total costs were estimated for these technologies providing a clear ranking with nuclear having the lowest costs and some of the renewables showing remarkable cost reductions until 2030. This ranking is partially controversial mainly due to the limited representation of social aspects in the total costs. The results of MCDA-applications involving elicitation of preferences from a relatively homogeneous stakeholder group, i.e. 85 employees of the Axpo Group (including also NOK, EGL, CKW and Axpo IT), are summarized. In addition, sensitivity of technology ranking to preference profiles is demonstrated. Broader consideration of social factors favours renewables and depending on the specifics of preference profiles may lower the ranking of nuclear. Further applications of the MCDA-approach with various stakeholder groups are planned. Ó 2008 Published by Elsevier Ltd.

1. Background One of the key issues of current and future energy policy in Switzerland is the ensuring of a sufficient electricity supply. Given the steadily growing electricity consumption (BFE, 2007), the forthcoming expiration of electricity import contracts with France and the need of replacement of some of the older Swiss nuclear plants, options for a sustainable electricity supply have to be considered and evaluated. Such an assessment should include all pillars of sustainability, i.e. it must cover environmental, economic and social aspects.

* Corresponding author. Tel.: +41 (0)56 310 2956; fax: +41 (0)56 310 4411. E-mail address: [email protected] (S. Hirschberg). 1 Present address: BKW FMB Energy Ltd., Berne, Switzerland. 0306-4549/$ - see front matter Ó 2008 Published by Elsevier Ltd. doi:10.1016/j.anucene.2008.11.029

In order to allow a transparent decision process, which could be accepted by all stakeholders, an interdisciplinary, comprehensive approach based on a multi-criteria decision analysis (MCDA) tool has been implemented. Based on this MCDA tool the sustainability of current and future power supply technologies relevant for Switzerland has been evaluated. This was accompanied by the assessment of total costs, covering the environmental, health and economic dimensions. 2. Objective and scope The goal of the study was to allow a complete and transparent comparison of all realistically available options for the Swiss electricity supply in the forthcoming years up to 2030 regarding their sustainability. For this purpose a tool has been developed, which builds on quantified indicators measuring the environmental,

410

S. Roth et al. / Annals of Nuclear Energy 36 (2009) 409–416

Table 1 Main environmental, economic and social criteria in the frame of the MCDA for sustainability assessment. Criteria

Responsibilities

Environment Resources Climate Change Impacts on ecosystems Waste

PSI PSI PSI PSI

Social aspects Physical security Political stability and legitimacy Social development Impacts on quality of landscape & residential areas Impacts on human health Social components of risks

PSI/UniStutt UniStutt UniStutt UniStutt PSI UniStutt

Economy Impacts on Impacts on Impacts on Impacts on

BAK/CEPE Axpo/PSI CEPE/Axpo/PSI Axpo/PSI/UniStutt

the GDP customers state affairs utility

economic and social performance of various power plants and their associated energy chains, and allows explicit allocation of weights or preferences to these indicators. The indicators and preferences are combined within the MCDA model for the calculation of the so-called ‘‘sustainability index”, which can be used for sustainability ranking of electricity produced by the different energy chains. The major Swiss energy supplier, Axpo, initiated and coordinated the project. Besides the Paul Scherrer Institut (PSI) participants in the project included also the University of Stuttgart (UniStutt), the Centre for Energy Policy and Economics (CEPE) at the Swiss Federal Institute of Technology Zurich (ETHZ), and BAK Basel Economics. The sustainability evaluation framework was defined in terms of a wide set of criteria, with the associated 75 indicators, quantified for the years 2000 and 2030. This includes 11 environmental, 33 social and 31 economic indicators. Not surprisingly the number of environmental indicators is lowest since in this case objective aggregation of environmental flows can be carried out based on natural sciences. This is mostly not possible for social indicators which constitute the largest group. To a large extent, current technologies were represented by the best commercially available options, while a wide spectrum of evolutionary concepts was included for the future technologies. Compared to earlier studies by PSI, which focused on power technology assessments in Switzerland (Hirschberg et al., 2000), Germany (Hirschberg et al., 2004a) and China (Eliasson and Lee, 2003), the present work embodies a broader set of technologies, and employs a much extended set of evaluation criteria and indicators. Table 1 shows the main environmental, social, and economic criteria along with the responsible institutes. Almost all of these criteria are further decomposed into sub-criteria not shown in the table.

BLW). PSI is responsible for inventories of all energy systems. For more details on the treatment of environmental indicators we refer to Bauer et al. (2007, 2008). The LCA methodology does not take into account site-specific effects of environmental burdens; however, this aspect is addressed within the social area through the quantification of impacts of air pollution on human health. Human health damages are quantified based on the state-of-the-art impact pathway approach, as defined in recent projects within the ExternE-series (Friedrich et al., 2004; Rabl and Spadaro, 2005). The analysis of accident risks is based to a large extent on the historical experience of accidents using ENSAD (energy-related severe accident database), developed by PSI. ENSAD is the most extensive and detailed database world-wide for severe accidents in the energy sector (Burgherr et al., 2004; Hirschberg et al., 2004b). The estimates for the year 2030 are supported by trend analysis of the historical data and extrapolations to the future. For hypothetical nuclear accidents probabilistic safety assessment (PSA) was employed building on simplified consequence assessment methodology earlier published in Hirschberg et al. (2003). Economic indicators build on extensive literature studies, a net present value model of the utility, inputs from industry, and, where appropriate, on expert judgement. Most of the social indicators under responsibility of the University of Stuttgart were quantified based on a survey among experts in the areas of energy technologies, sustainability, and energy policy as well as risk analysis. This survey included a final Delphi application among the experts. In relative terms the social indicators and some economic indicators due to their nature exhibit a higher degree of subjectivity than the environmental ones. 4. Reference technologies The technology portfolio contains both large centralized power plants and smaller decentralized units in Switzerland and few other European countries (for electricity imports). Small combined heat and power units burning natural gas or gasified biomass were assessed along with base- and mid-load large power plants. Evolutionary technology development was assumed to take place between today and 2030 for all reference power plants. For most of the analysed technologies specific locations in Switzerland were assumed. For some centralized fossil technologies as well as for nuclear and wind power plants also locations in other countries were considered. For hydro NOK-plants were selected as reference. They may not be generally representative for hydro power plants in Switzerland. Tables 2 and 3 summarize the most important characteristics of the selected reference technologies. For more details on the reference technologies we refer to Bauer et al. (2008).

5. Results

3. Quantification methods used

5.1. Total costs

The quantification of ecological indicators is based on life cycle assessment (LCA) (Dones et al., 2007), which means that full energy chains, extraction and processing of resources, manufacturing of products and infrastructure, transport of materials, conversion of fuels, and waste disposal are included in the analysis. The background LCA database was ecoinvent data v1.2, reflecting conditions in year 2000. Ecoinvent (ecoinvent, 2008) is a centralized, webbased LCA database developed and implemented by the Swiss Centre for Life Cycle Inventories (EPFL, EMPA, ETHZ, FAL, PSI) and supported by several Swiss Federal Offices (BUWAL, BfE, ASTRA, BBL,

Fig. 1 shows the estimated total costs of electricity supply (internal production costs plus external costs resulting from health and environmental damage), which have been proposed as a possible aggregate measure of sustainability. The results are provided for time frames around years 2000 and 2030. Nuclear power has the lowest total costs (both current and in 2030), followed by hard coal and natural gas power plants, runof-river hydro power and wind power at optimal sites. Remarkable cost improvements are expected for new renewables with biogas and wind onshore in Germany reaching competitive production

Table 2 Main characteristics of current reference technologies. Nuclear

Nuclear

Hard coal

Natural gas

Natural gas

Natural gas

Natural gas

Natural gas

Hydro power

Technology

Pressure water reactor, Generation II 730 –

Pressure water reactor, Generation II 1300 –

Supercritical steam cycle (SC), base load 509 –

Combined cycle (CC), base load 400 –

Combined cycle (CC), mid load 400 –

Combined cycle (CC), base load 400 –

Combined heat & power (CHP) 0.2 0.3

Solid oxide fuel cell (SOFC) 0.2 0.2

Run-of-river 51 –

France (F), Cattenom

Germany (D), Rostock

Switzerland, Birr

Switzerland, Birr

Italy (I), Naples

Switzerland, Baden

Operating time (full load hours per year) Efficiency electric (%) Lifetime (a)

Switzerland (CH), Beznau 8000

6300

7000

8000

4000

8000

4500

Switzerland, Baden 4500

Switzerland, Wildegg-Brugg 5720

32.0 4C

34.0 4C

43.2 3C

57.5 25

55.5 25

55.5 25

32.0 2C

40.0 5

88.9 80

Energy source

Hydro power

Biogas

Synthetic natural gas (SNG)

Wind power

Wind power

Wind power

Photovoltaic

Photovoltaic

Geo thermal

Technology

Reservoir

Combined heat & power (CHP)

Combined heat & power (CHP)

Onshore wind park, 4 turbines

Onshore wind park, 50 turbines

Offshore wind park, 80 turbines

Multicrystalline-Si panel, roof-top

Amorphous-Si, roof-top

Capacity el. (MWel) Capacity fh. (CHP) (MWth) Location

53 –

0.1 0.1

0.2 0.3

4  0.85 –

50  2 –

80  2 –

0.02 –

0.01 –

Enhanced Geo thermal System (EGS) 3 –

Switzerland, Ilanz/ Panix 2476

Switzerland, Baden

Switzerland, Baden

Denmark (DK), Horns Rev 3750

850

Switzerland, Baden 850

Switzerland, Basel

4500

Germany (D), North Sea coast 2500

Switzerland, Baden

7000

Switzerland, ML Crosin 1250

89.0 150

36.0 15

32.0 20

n.s. 20

n.s. 20

n.s. 20

14.4a 30

6.5b 30

11.3 3C

Capacity el. (MWel) Capacity fh. (CHP) (MWth) Location

Operating time (full load hours per year) Efficiency electric (%) Lifetime (a) a b

S. Roth et al. / Annals of Nuclear Energy 36 (2009) 409–416

Energy source

7000

Cell efficiency; module efficiency is 13.2%. Average cell efficiency over lifetime taking into account degradation.

411

412

Table 3 Main characteristics of future (year 2030) reference technologies. Nuclear

Nuclear

Hard coal

Natural gas

Natural gas

Natural gas

Natural gas

Natural gas

Hydro power

Technology

European pressure water reactor (EPR), Generation III

European pressure water reactor (EPR), Generation III

Integrated gasification combined cycle (IGCC)

Combined cycle (CC), base load

Combined cycle (CC), mid load

Combined cycle (CC), base load

Combined heat & power (CHP)

Run-of-river

Capacity el. (MWel) Capacity fh. (CHP) (MWth) Location

1500 –

1500 –

450 –

500 –

500 –

500 –

0.2 0.21

Solid oxide fuel cell (SOFC) 0.2 0.11

51 –

Switzerland (CH), Beznau

France (F), Cattenom

Germany (D), Rostock

Switzerland, Birr

Italy (I), Naples

Operating time (full load hours per year) Efficiency electric (%) Lifetime (a)

8000

8000

7000

8000

Switzerland, Birr 4000

8000

Switzerland, Baden 4500

Switzerland, Baden 4500

Switzerland, Wildegg-Brugg 5720

33.8 60

33.8 60

51.5 30

63 25

61 25

61 25

42 20

52 15

88.9 80

Energy source

Hydro power

Biogas

Synthetic natural gas (SNG)

Wind power

Wind power

Wind power

Photovoltaic

Photovoltaic

Geothermal

Technology

Reservoir

Combined heat & power (CHP)

Combined heat & power (CHP)

0.2 0.15

0.2 0.21

Offshore wind park, 50 turbines 80  20 –

AmorphousSi, roof-top

53 –

Onshore wind park, 50 turbines 50  4.5 –

Multi crystal lineSi panel, roof-top

Capacity el. (MWel) Capacity th. (CHP) [MWth| Location

Onshore wind park, five turbines 42 –

0.02 –

0.01 –

Enhanced geothermal system (EGS) 36 –

Switzerland, Ilanz/Panix

Switzerland, Baden

Switzerland, Baden

2476

7500

4500

Germany, North Sea coast 2700

Denmark (DK), North Sea 4000

Switzerland, Baden 850

Switzerland, Baden 850

Switzerland, Basel

Operating time (full load hours per year) Efficiency electric (%) Lifetime (a)

Switzerland, Mt Crosin 1500

89.0 150

41.7 15

42 20

n.s. 20

n.s. 20

n.s. 20

20.0a 30

13.7b 30

11.3 3C

a b

Cell efficiency; module efficiency is 19%. Average cell efficiency over lifetime taking into account degradation.

7000

S. Roth et al. / Annals of Nuclear Energy 36 (2009) 409–416

Energy source

S. Roth et al. / Annals of Nuclear Energy 36 (2009) 409–416

413

Fig. 1. Full costs of electricity generation options around years 2000 and 2030 (Hirschberg et al., 2008); not all of the options analyzed are shown. (Gen II/III = Generation II and III reactors; SC = supercritical; IGCC = integrated gasification combined cycle; CC = combined cycle; CHP = combined heat and power; SOFC = solid oxide fuel cell.)

cost levels, further strengthened by the very low external costs. Electricity from fuel cells and photovoltaic modules is today at the other end of the cost range. Even if technological progress is expected to allow drastic cost reduction for these systems, their relative ranking in terms of total costs is not expected to change within the considered time period. New renewables in general will very likely profit from the most significant cost reduction in the next decades. Some of them will be competitive with coal and gas power plants, also because fossilbased electricity production will become more expensive. In fact, due to recent fuel price increases the current costs of gas-based electricity are substantially higher than those shown in the figure. It should be noted that the potential of new renewable electricity supply options in Switzerland is quite limited in the mid-term (Hirschberg et al., 2005). Consideration of external costs generally improves the competitiveness of the renewables and nuclear power options in comparison to those based on fossil fuels. External costs of the selected high performance fossil-fuel technologies are dominated by the damage caused by greenhouse gas (GHG) emissions leading to climate change followed by impacts due to major air pollutants. In relative terms the estimates associated with GHGs are much more uncertain and depending on the assumptions could be much higher. The total cost approach is very useful for carrying out cost-benefit analyses, but its use in the assessment of the relative sustainability of the various options is not fully accepted. The main objections derive not only from limited coverage of social aspects but also from the questionable acceptability of monetary values being applied to a few of the social indicators that have been explicitly addressed. Consequently, MCDA has been employed to provide an alternative quantification of an aggregated sustainability indicator. 5.2. Multi-criteria decision analysis The complete MCDA model, with some of the criteria not readily amenable to monetization (e.g. risk perception and political stability), does allow for the explicit consideration of individual or

group preferences through assignment of appropriate weighting factors and therefore has a better chance to be accepted by a wider range of stakeholders. These preferences are combined with technology-specific indicators. The approach used for the evaluation is based on a simple weighted multiple attribute function. The weights reflect the relative importance of the various evaluation criteria and are combined with the normalized indicator values (scores). Normalization is carried out using a local scale, defined by the set of alternatives under consideration. For example, the alternative which does best on a particular criterion is assigned a score of 100 and the one which does least well a score of 0; based on linear interpolation all other alternatives are given intermediate scores which reflect their performance relative to these two end points. A single overall value is obtained for each alternative by summing the weighted scores for all criteria. Ranking of the available options is then established on the basis of these values. As opposed to the total costs aggregation MCDA provides no fixed ranking of technologies. The results may change depending on trade-offs expressed through specific preference profiles of the various stakeholders. The weights can be obtained from stakeholders. Alternatively, various weighting schemes can be assigned to accommodate a range of perspectives expressed in the energy debate. The sensitivity to these schemes can be investigated. An interactive MCDA tool has been developed and used in a series of workshops. In the following examples of MCDA results are provided. This includes results obtained in the workshops attended by selected employees of the Axpo Group as well as examples utilizing preference profiles constructed by the authors. Figs. 2–4 summarize the results from the workshops, based on preferences expressed by 85 employees of the Axpo Group. The participants in the exercise may be seen as representing a specific stakeholder group though their preferences reflect personal opinions and exhibit quite large variation. Within the procedure used at the workshop the participants were first asked to provide their intuitive ranking of the technologies of interest. This was followed by the allocation of individual preferences to indicators selected by them (maximum 30 with at

Average Ranking 18 16 14 12 10 8 6 4 2 0 11.6 11.7 10.8 10.9

8 15.1 17

8.6 8.1 8.4 8.5

4.5

1.8 3.1 3.2

Fig. 3. Average ranking of technologies by Axpo-participants in MCDA-workshops.

least three within each of the three dimensions of sustainability), leading to the establishment of ranking based on MCDA. Each participant could then change the preferences, if desired. Finally, the individual MCDA results could be compared with the intuitive ranking and main factors driving the results could be identified. Fig. 2 shows the average preferences and indicator selections of the participants for the whole spectrum of criteria and indicators used. The three dimensions of sustainability were selected by all participants and weighted on average in a very balanced way, i.e. environment at 36%, social at 29% and economy at 35%. The highest weights for individual indicators were assigned to greenhouse gas (GHG) emissions, consumption of fossil resources and electricity prices. Very low or relatively low weights were on average assigned to such aspects as consumption of uranium resources and minerals, acidification and eutrophication, nuclear proliferation, Initial estimation of ranking vs. MCDA result (>0 better than estimation)

13.6 13.6 12.6

Na t. ga s NP CH P, P Ha rd NP CH P N c Na at. oal, , F t. ga GE g s Hy as SO R dr CC FC o: , re bl. Hy s , d Na ro PV erv I t. : ru , m oir ga n c s -of -S CC -r i Na , b iver l. t. ga PV , CH s -, CC aS Bi , m i og a l., W in SN s C I do G H W ns C P in h. H W doff , GE P i G ndo sh., R eo n th sh DK er ., m CH al HD R

yd ro H : ru yd nG ro of eo : r -ri th es ve er er r m v Bi al oir og H a D W N sC R in P H d P P on , C s H W S h ., in N C N Win d o G C H at d ff H . g o sh P as ns ., C h., D K C G ,m E N id R at lo .g as P NP ad P N V C at C ,m ,F .g ,b c as as -S e i C C P loa ,b V d N as , a at e -S . g lo i N as ad, a H t. g S O I ar a F d s C co C a l HP ,G ER

H

Indicator chosen

50%

90% 45%

80% 40%

70% 35%

60% 30%

50% 25%

40% 20%

30% 15%

20% 10%

10% 5%

0% 0%

8

6

-8

Average Weight

100%

ENVIRONMENT RESOURCES Fossil Energy Uranium Metals CLIMATE CHANGE ECOSYSTEM QUALITY Land use Ecotoxicity Acidification & eutrophication Land contamination WASTE Non-radioactive Radioactive SOCIAL ASPECTS INNER SECURITY Terrorist threat Max. number of fatalities Loss of production Cost of reconstruction Availability of disposal infrastructure Availability of disposal concept POLITICAL STABILITY & LEGITIMACY Potential of conflicts Potential of mobilisation Post operational safeguarding Proliferation Conflicts over resources Controllability of conflicts Existence of conflict resolution mechanisms Social cooperation Trust in utility Qualitative risk characteristics Participation of residents SOCIALLY ACCEPTABLE DEVELOPMENT Economic development of site region Socio-economic image Impacts on local infrastructure Satisfaction of residents Equity Fair distribution of risks & benefits Electricity for economically weak groups SITE QUALITY Quality of living conditions Noise impacts on residents Site dependent traffic Benefits for regional sustainability Impulses for sustainable utility behaviour Impulses for sustainable consumer behaviour Impacts on quality of landscape Direct land use Aestetic impacts IMPACTS ON HUMAN HEALTH Normal operation Mortality Morbidity Severe accidents Fatalities Injured Evacuees SOCIAL COMPONENTS OF RISKS Perceived health risks (normal operation) Accident risks Perceived health risks (accidents) Perceived safety management competence Overexploitation of renewable resources ECONOMY CONTRIBUTION TO THE GDP Contribution to the national economy Employment Jobs in alpine regions Jobs in non-alpine regions New jobs in non-alpine regions Qualification of employees Innovation Education of employees Jobs in R&D Technology transfer Development of new products/services EFFECTS ON CUSTOMERS Effect on electricity cost CONTRIBUTION TO GOVERNMENTAL DUTIES Autonomy of electricity production Cash flow to the state External costs & benefits EFFECTS ON OPERATOR Profits Financial risks Volatility of fuel costs Risks due to authorities' interventions Necessary measures in advance & after operation Operator flexibility Liquidity Time for construction of the plant Flexibility based on marginal costs Flexibility of production Limitations in electricity production Predictability of energy availability Technical site availability Impacts on image of operator Compatibility with Axpo's corporate culture

414 S. Roth et al. / Annals of Nuclear Energy 36 (2009) 409–416

Fig. 2. Average preferences and indicators selected by Axpo-participants in MCDA-workshops.

12

10

subjectively overestimated technologies

4

2

-2 0

-4

-6

subjectively underestimated technologies

-10

Fig. 4. Discrepancies between intuitive and MCDA-based rankings by Axpoparticipants in MCDA-workshops.

mobilization potential against specific options, human health impacts of normal operation and accidents but also unexpectedly to most indicators expressing the impact on the utility. In Fig. 3 the average ranking of each of the 18 technologies is shown. Top positions are taken by both hydro options, followed by geothermal, biogas, nuclear in Switzerland, wind options and synthetic natural gas from wood (SNG). Somewhat surprisingly wind in Switzerland ranks slightly better than wind in Germany and Denmark, where conditions are superior. Combined cycle

415

S. Roth et al. / Annals of Nuclear Energy 36 (2009) 409–416

(CC) natural gas ranks at about the same level as nuclear in France and solar PV. Lowest rankings are obtained for cogeneration based on natural gas and coal in Germany. It may be worthwhile to consider the ranking of centralized and decentralized technologies separately given the very large differences in their technical and economic potentials at least in the mid-term perspective. Fig. 4 shows the discrepancies between the intuitive and MCDAbased rankings. The differences are on average relatively small with practically no discrepancies for hydro, CC and solar PV. In relative terms gas-based cogeneration and nuclear have been mostly overestimated by intuitive judgements while geothermal and wind options belong to the most underestimated. The results based on average preferences of a specific stakeholder group do not reflect the influence of various preference pro-

files on the technology ranking. This effect is illustrated here by using two such profiles constructed by the authors. Both of them are balanced in the sense that the environmental, social and economic components of sustainability are equally weighted at the highest level of the hierarchy of criteria. Otherwise they strongly differ. The first set of preferences puts in relative terms more emphasis on the protection of climate and human health as well as on electricity generation costs and impacts on the utility. The highest weights are in this case assigned to GHG-emissions, electricity production costs, and mortality and morbidity associated with the normal operation. The second profile, while still having substantial but lower weights on GHGs and costs, emphasizes more resource depletion including uranium, internal security, political stability and legitimacy including proliferation potential,

Sustainability indicator [points]

100 90 83 80

77

79

79

77

76

74 70

70 63

64

73

65

63

57

60

54 50

56

50

50 40 30 20 10

N PP ,C H N H PP yd ro ,F H yd : r e se ro :r rv u oi r H n-o N a at . g rd f-riv c as e o N C al, r at C G . , E N a t gas b a s R .g el C oa as C d C ,m C i d ,b lo as ad el oa Bi d, og I as N C at H .g P as C SN H G P N at . g CH as P S PV OF ,m C cS W PV, i in a d -S o i W in nsh d . , on C W sh. H in ,G d E o G eo f f s h R th ., er D K m al H D R

0

Fig. 5. Technology ranking based on a constructed preference profile with equal weighting of the three dimensions of sustainability and in relation to the profile behind the results shown in Fig. 6 stronger emphasis on climate and human health protection, electricity generation costs and impacts on utility.

Sustainability indicator [points]

100 90 80 73

76

76

74 68

70 60 60

61 57

62

60

65 60

56

56

66

65

58

50 50 40 30 20 10

N PP ,C H N H P yd P ro ,F H yd : r e s ro : r erv o un H -o i r ar N f-r d at iv c .g a s oal er , N at CC GE N . ga , ba R at . g s C sel o C as , m ad C id C lo ,b as ad e Bi og loa as d , N C I at . g HP as S N CH G P N at . g CH as P S PV OF ,m C cS W PV, i in a d -S o i W in nsh d . , on C W sh. H in ,G d E o G eo ffsh R th ., er D K m al H D R

0

Fig. 6. Technology ranking based on a constructed preference profile with equal weighting of the three dimensions of sustainability and in relation to the profile behind the results shown in Fig. 5 stronger emphasis on resource depletion including uranium, internal security, political stability and legitimacy including proliferation potential, risk perception and consequences of severe accidents, and direct employment effects.

416

S. Roth et al. / Annals of Nuclear Energy 36 (2009) 409–416

risk perception and consequences of severe accidents, and direct employment effects. The first profile is likely to be closer to the views of the electrical industry while the second profile to those of major NGOs. Figs. 5 and 6 show the results obtained using the two profiles. In the first case hydro, geothermal, nuclear and biogas and wind options belong to the top performers. Next follow CC-systems and SNG; solar PV, SOFC, gas cogeneration and coal are the worst performers based on this profile. For the second profile hydro, geothermal and biogas remain on top, followed by SNG and wind while nuclear performs worse than in the first case, i.e. now comparably to solar PV. Using the second preference profile the differences in scores between the various technologies become smaller. 6. Conclusions The work reported in the present paper provides an extensive set of indicators characterizing the environmental, social and economic performance of current and future electricity supply technologies. These indicators can be used individually for comparative purposes but can also be aggregated partially by means of total costs or more extensively using MCDA. Rankings of technologies based on MCDA differ from those based on total costs. The latter provides a definite ranking while MCDA gives no single answer since there is a clear dependence of ranking on the assigned preference profiles. The much broader consideration of social factors as well as inclusion of issues related to the national economy in MCDA, tends to favour renewables while the impact on the ranking of nuclear is that it becomes lower. This emphasizes the desirability of nuclear technology developments towards practically excluding the possibility of large radioactivity releases and reducing the necessary confinement time of nuclear wastes to few hundred years. Such advancements are currently pursued within some of the Generation IV concepts. The MCDA model approach allows to carry out the assessment in a transparent manner by combining indicators that to the extent possible are based on recognized scientific approaches with individual or group preferences. In this way MCDA may serve the purpose of supporting and guiding the debate on the current and future energy supply. The implementation reported in the current paper will be subject to further extensions. At this stage only individual electricity supply technologies have been analysed. Of interest is to consider the whole supply system, i.e. technology mixes, taking into account relevant constraints, dependencies and interactions. Furthermore, it is desirable to engage a much wider spectrum of stakeholders. This is currently being done within the Swiss Energy Trialog process. In parallel PSI leads a technology road map and stakeholder

perspectives study for the EU, conducted within the NEEDS project. A set of criteria and indicator has been established (Hirschberg et al., 2007) and found acceptable by stake holders participating in an European survey (Burgherr et al., 2008). Web-based MCDA implementation for France, Germany, Italy and Switzerland is currently in progress. References Bauer, C., Dones, R., Heck, T., 2007. Comparative environmental assessment of current and future electricity supply technologies for Switzerland. In: Third International Conference on Life Cycle Management, 27–29 August 2007, Zurich, Switzerland (proceedings available on CD). Bauer, C., Dones, R., Heck, T., Hirschberg, S., 2008. Environmental assessment of current and future Swiss electricity supply options. In: International Conference on Reactor Physics, Nuclear Power: A Sustainable Resource, September 14–19, 2008, Interlaken, Switzerland. BFE, 2007. Gesamtenergiestatistik 2006. Burgherr, P., et al., 2004. Severe accidents in the energy sector. PSI Report for European Commission Within Project NewExt on New Elements for the Assessment of External Costs From Energy technologies. Paul Scherrer Institute, Villigen, Switzerland. Burgherr, P., Hirschberg, S., Schenler, W., 2008. Survey II on sustainability criteria and indicators: approach and results. Report of Research Stream 2b, EU project NEEDS. Dones, R., Bauer, C., Bolliger, R., Burger, B., Faist Emmenegger, M., Frischknecht, R., Heck, T., Jungbluth, N., Tuchschmid, M., Röder, A., 2007. Sachbilanzen von Energie systemen: Grundlagen für den ökologischen Vergleich von Energiesystemen und den Einbezug von Energiesystemen in Öko bilanzen für die Schweiz. Final report ecoinvent v2.0 No. 6. Paul Scherrer Institut Villigen and Swiss Centre for Life Cycle Inventories, Dübendorf, Switzerland. ecoinvent, 2008. Website of the ecoinvent database: . Eliasson, B., Lee, Y. (Eds.), 2003. Integrated Assessment of Sustainable Energy Systems in China. The China Energy Technology Program. Friedrich, R., et al., 2004. New Elements for the Assessment of External Costs From Energy Technologies. EU 5th Framework Programme. Hirschberg, S., et al., 2000. Use of external cost assessment and multi-criteria decision analysis for comparative evaluation of options for electricity supply. In: Kondo, S., Furuta, K., (Eds.) PSAM5, 27 November–1 December, 2000, Osaka, Tokyo, pp. 289–296. Hirschberg, S. et al., 2003. In: Assessment of severe accident risks. In: Eliasson, B., Lee, Y.Y. (Eds.), . Integrated Assessment of Sustainable Energy Systems in China – The China Energy Technology Program. Kluwer Academic Publishers, Dordrecht, Boston, London, pp. 587–660. Hirschberg, S., et al., 2004a. Sustainability of electricity supply technologies under German conditions: a comparative evaluation. PSI Report No. 04-15. Hirschberg, S. et al., 2004b. Severe accidents in the energy sector: comparative perspective. Journal of Hazardous Materials 111, 57–65. Hirschberg, S., et al., 2005. Neue Erneuerbare Energien und neue Nuklearanlagen: Potenziale und Kosten. PSI-Report No. 05-04. Paul Scherrer Institut, Villigen, Switzerland. Hirschberg, S., Bauer, C., Burgherr, P., Dones, R., Schenler, W., Bachmann, T., Gallego Carrera, D., 2007. Environmental, economic and social criteria and indicators for sustainability assessment of energy technologies. Deliverable n° D3.1 – Research Stream 2b, EU Project NEEDS. Hirschberg, S., Bauer, C., Burgherr, P., Cazzoli, E., Dones, R., Heck, T., Schenler, W., 2008. Treatment of risks in sustainability assessment of energy systems. In: Proceedings of PSAM 9, 9th International Probabilistic Safety Assessment & Management Conference, 18–23 May 2008, Hong Kong, China. Rabl, A., Spadaro, J. V. (Eds.), 2005. Externalities of Energy: Extension of Accounting Framework and Policy Applications. Version 2, EC EESD-Programme.