Renewable Energy 31 (2006) 5571 www.elsevier.

com/locate/renene

Dynamic life cycle assessment (LCA) of renewable energy technologies

Martin Pehnt*
Institute for Energy and Environmental Research Heidelberg (Ifeu), Wilckensstr. 3, D-69120 Heidelberg, Germany Received 7 October 2004; accepted 5 March 2005 Available online 11 May 2005

Abstract Before new technologies enter the market, their environmental superiority over competing options must be asserted based on a life cycle approach. However, when applying the prevailing status-quo Life Cycle Assessment (LCA) approach to future renewable energy systems, one does not distinguish between impacts which are imported into the system due to the background system (e.g. due to supply of materials or nal energy for the production of the energy system), and what is the improvement potential of these technologies compared to competitors (e.g. due to process and system innovations or diffusion effects). This paper investigates a dynamic approach towards the LCA of renewable energy technologies and proves that for all renewable energy chains, the inputs of nite energy resources and emissions of greenhouse gases are extremely low compared with the conventional system. With regard to the other environmental impacts the ndings do not reveal any clear verdict for or against renewable energies. Future development will enable a further reduction of environmental impacts of renewable energy systems. Different factors are responsible for this development, such as progress with respect to technical parameters of energy converters, in particular, improved efciency; emissions characteristics; increased lifetime, etc.; advances with regard to the production process of energy converters and fuels; and advances with regard to external services originating from conventional energy and transport systems, for instance, improved electricity or process heat supply for system production and ecologically optimized transport systems for fuel transportation.

* Tel.: C49 6221 4767 36; fax: C49 6221 4767 19. E-mail address: [email protected]. URL: www.ifeu.de.

0960-1481/$ - see front matter q 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2005.03.002

56

M. Pehnt / Renewable Energy 31 (2006) 5571

The application of renewable energy sources might modify not only the background system, but also further downstream aspects, such as consumer behavior. This effect is, however, strongly context and technology dependent. q 2005 Elsevier Ltd. All rights reserved.
Keywords: LCA; Life cycle assessment; Forecasting; Renewable energy; Photovoltaics; Geothermal; Hydropower; Biomass; Wind; Solar thermal

1. Introduction Technological advances in the eld of distributed and renewable energy systems, the requirement of climate gas mitigation and electricity system capacity decits, but also market restructuring and deregulation have led to an increasing interest in innovative energy technologies. Before new technologies enter the market, however, their environmental superiority over competing options must be asserted based on a life cycle approach. Life Cycle Assessment (LCA) investigates environmental impacts of e.g. systems or products from cradle to grave throughout the full life cycle, from the exploration and supply of materials and fuels, to the production and operation of the investigated objects, to their disposal/recycling. With the increasing environmental operation standards of modern energy conversion systems, the upstream and downstream processes, e.g. fuel supply or power plant and infrastructure production, become increasingly relevant [1]. In the prevailing status-quo LCA approach, future developments of the energy systems themselves and of the context in which the systems are to be applied are typically not considered, thus severely distorting the analysis of the environmental characteristics of future energy systems. In a causal dimension, the following questions arise: Which of these environmental impacts can be causally attributed to renewable energies (inherent impacts), and which are imported into the system due to the background system?1 What is the improvement potential of these technologies compared to that of competitors technologies, e.g. due to process and system innovations or diffusion effects (e.g. ecology of scale: lower production impacts due to higher sales numbers) [1]? These questions also lead to a time dimension: How fast will the background system change? How fast will the improvement potentials be made accessible?
In LCAs, background systems are system components that are not directly part of the product systems but which are necessary for the production, use, and disposal of these, e.g. the electricity supply mix for the production of a power plant or the transport infrastructure for fuel transport.
1

M. Pehnt / Renewable Energy 31 (2006) 5571

57

Using a dynamic2 rather than a static approach helps to identify the inherent environmental bottlenecks. For instance, today under German conditions, producing a polycrystalline solar-grade Silicon photovoltaics system leads to greenhouse gas emissions of 100 g CO2 equiv./kW hel. From these, a large part is imported into the product system, e.g. because fossil energy is used within the production process. Taking into consideration a future energy mix for production, higher recycling rates, advances with respect to wafer losses, module efciencies, and a higher lifetime cuts the emissions to approximately 50 g CO2 equiv./kW hel. This paper investigates the environmental performance of renewable energy systems particularly in view of future developments.

2. First step: static LCA of renewable energy technologies 2.1. Methodology, goal and scope The rst step of this exercise is to set up LCA models of the respective status-quo renewable energy systems. For this purpose, networks in the LCA software package Umberto (www.umberto.de), which are the basis for life cycle inventory and impacts assessment, are set up. The LCA results are analyzed with regard to critical life cycle segments and materials and compared to conventional systems. To this end, data from manufacturers and system operators is compiled and the extensive IFEU database used, complemented with data from various literature LCAs (wind power [2], solar thermal power plants [3], geothermal energy [4], PV [5], solar thermal collectors [6], biogas [7]). The materials, energy supply chains, transport services, etc. are modeled with the Umberto database (www.umberto.de). A more precise denition can be found in [8]. The functional unit used in the system of electricity generation described in this paper is one 1 kW hel at the power plant3 for the electricity generating system and 1 kW h at the heat distribution system in a house for the heat generating system. The geographic reference for the assessment of renewable energy technologies is the Federal Republic of Germany; the time reference is 2010. The most recent LCA data was taken for the assessment. If signicant changes are to be expected until 2010, the data is adapted for the general conditions in 2010. Processes assessed are production, operation and maintenance, and system recycling/ disposal. The infrastructure of supply of fuels and power plants was considered with the exception of the utilization of roads due to lorry transports. Unless stated otherwise, recycling is assessed for a closed loop recycling, i.e. it is assumed that recycled material can substitute the use of the primary material to a certain percentage. The expenses of
2 In this context, dynamic does not necessarily mean that the development of the product and background system is modelled continuously, but rather it means that a future state of the system is modelled considering the future characteristics of the background and the model system. 3 This system boundary was chosen deliberately, because the electricity distribution is characterized by signicant data uncertainty, particularly with respect to avoided or extra losses due to distributed energy systems and with respect to material input for the electricity grid.

58

M. Pehnt / Renewable Energy 31 (2006) 5571

Table 1 Considered impact categories and characterization factors U in streamlined LCAs Impact category Energy resources Global warming Inventory parameter CO2 CH4 N2O SO2 NOx NH3 HCl NOx NH3
a

Characterization factor U CED Global warming potentiala

Reference MJ CO2 equiv. CO2 equiv. CO2 equiv. SO2 equiv. SO2 equiv. SO2 equiv. SO2 equiv.

Value U (kgMaterial/ kgReference) 1 21 310 1 0.7 1.88 0.88 0.13 0.33

Acidication

Acidication potential

Eutrophication

Eutrophication potential

K PO3 4 equiv. K PO3 4 equiv.

Time horizon 100 years.

recycling material processing are allocated to the process. Necessary allocation or credit is described in the respective sections about the technology. The impact categories include energy resource consumption (also called simplied cumulated energy demand), non-energy resource consumption, and emission of greenhouse gases, eutrophication, and acidication. The characterization factors are summarized in Table 1. Due to the streamlined character of the LCA, only a limited number of inventory parameters are assessed here. However, for all technologies it was checked whether there are specic substances involved that would need to be taken into consideration (e.g. in magnesium production, SF6 is emitted. If magnesium were involved in any of the systems, the signicance of SF6 to total global warming was checked). The impact category of land use is not documented. This was considered in greater detail by means of geographic information systems in [8] and will be reported elsewhere. Finally, the results are normalized. The normalization takes place for electricity generating systems with regard to electricity mix for Germany in 2010 (Table 2). That is, impacts of provision of 1 kW hel by means of renewable energy systems are divided by the impacts of the assumed electricity mix as dened in the business-as-usual development
Table 2 Environmental impacts of the future German electricity and heat mix Electricity mix 2010 per kW hel Iron ore (Finite) energy resources Global warming Acidication Eutrophication g MJ g CO2 equiv. mg SO2 equiv. K mg PO3 4 equiv. 2.6 8.91 566 1083 59.9 Heat mix 2010 per MJth 0.2 1.23 81.5 115 7.7

These factors are used for the normalization.

M. Pehnt / Renewable Energy 31 (2006) 5571

59

(energy carrier split and average power plant efciency) according to the reference scenario of the German Enquete commission [9]. In other words, a value higher than 100% implies that in the relevant environmental impacts there is an increase in detrimental effect in comparison to the mix; a value below 100% means a reduction. This normalization serves two purposes. On the one hand, environmental advantages and disadvantages of the electricity consumption can be identied easily. On the other hand, different environmental impacts can be represented in one diagram. The heating systems are normalized to a heuristic heat mix of 54% natural gas condensing boilers and 46% oil boilers, thus representing the present ratio for oil and gas heating (Table 2). 2.2. Results The results for selected streamlined LCAs of electricity and heat producing systems are presented in Fig. 1. The results of the inventory and impact assessment are presented in Tables 3 and 4. Greenhouse gas emissions and the consumption of non-renewable energy resources of renewable energy systems are signicantly lower compared to those of conventional systems. The electricity values have a maximum of 20% of the 2010 electricity mix and heat values have a maximum of 15% of the heat mix. In the case of biomass systems obtaining heat and electricity credits, a negative environmental effect arises depending on the system type, i.e. the substitution effect results in the environmental relief for the entire system. With regard to material resources (iron ore, bauxite), a smaller or similar impact arises as does in the case of conventional systems. Exceptions include photovoltaics due to mounting the modules, solar collectors due to aluminum consumption for the collectors
Iron ore requirements Finite energy resources Greenhouse effect Acidification

Power and/or heat mix 2010 = 100%

250%

200%

150%

Technology status 2010 Hydro: small run-of-river PV: 3 kW with polycrystalline solar silicon Wind: onshore 1.5 MW Solar thermal: parabolic trough Geothermal: Hot Dry Rock Wood PP: Steam turbine power plant with forest wood/short rotation forestry (SRF) Solar collector: local heating system with long-term heat storage Straw: straw-fired heating plant Wood: central heating with forest wood

490 %

100%

50%

0% Hydro PV Wind Solar Geothermal Wood PP SRF thermal Forest Power Systems Wood Solar Straw collector Heating Systems

Fig. 1. Normalized LCA of selected renewable energy systems for selected impact categories (full results see Tables 3 and 4).

Table 3 Selected inventory and impact assessment results of renewable electricity systems
Product Unit HydroHydropower power 3.1 MWel 300 kWel 1 kW hel Wind 1.5 MW (Onshore) Wind 2.5 MW (Offshore) Geother- Solar therPV mal (Hot mal (Para(polyc. bolic SOG-Si) Dry trough Rock 80 MWel) Forest wood steam turbinea SRF steam turbinea Waste wood steam a,b turbine SRF coForest combuswood Co-com- tion bustion Forest wood reciprocating enginea 1 kW hel and 1.7 kW hth 0.36 1.5 93 27 77 29 74 829 1360 157 87 0.2 0.1 0.5 272 SRF reci- Biogasa procating enginea

60

1 kW hel 1 kW hel 1 kW hel 1 kW hel 1 kW hel 1 kW hel

1 kW hel 1 kW hel 1 kW hel 1 kW hel 1 kW hel

1 kW hel and 1.7 kW hth 0.53 3.5 81 41 124 130 111 898 1349 149 125 1 137 0.4 489

1 kW hel and 0.39 kW hth

Ressources CED MJ Iron ore g Bauxite mg Emissions in air CO2 g CH4 mg N2O mg mg SO2 CO mg NOx mg mg NMHCc mg Particles/ dust HCl mg NH3 mg Benzene mg Benzo(a)- mg pyrene Impact assessment g Global warming Acidifcamg tion Eutrophi- mg cation

M. Pehnt / Renewable Energy 31 (2006) 5571

0.10 1.7 4 10 21 0.4 17 59 36 6 26 0.1 0.04 0.03 0.2

0.14 2.0 16 13 29 0,7 28 74 49 11 31 0.2 0.06 0.05 0.3

0.12 3.3 4.8 10.2 24.1 0.2 39.5 96.8 31.1 26.1 42.2 0.2 0.03 0.02 0.48

0.11 5.1

1.5 3.3 1200 99 220 1.9 288 141 340 20 119 2.9 0.71 0.09 1.4

0.54 3.2 4.7 37.8 103.4 2.6 61.6 208 188.9 35.4 1.1 0.7 0.05 0.3

0.14 2.78 7.15 13.4 35.2 0.2 46.7 85.4 72.9 2.1 40.1 0.4 0.14 0.22 0.36

0.28 1.0 29 22 17 73 72 757 1064 45 60 41 0.1 2.7 251

0.46 2.8 20 35 58 161 198 820 1192 40 95 42 119 2.6 447

0.36 3.7 27 31 63 14 315 405 1320 123 109 55 0.1 44.9 502

0.18 0.7 19 14 21 41 26 185 258 30 86 5 14 2.1 122

0.29 1.8 13 23 47 98 67 226 330 27 109 5 91 2.0 248

0.09 2.5 34 11 K19,763 K743 368 723 575 166 38 0.1 1619 0.02 0.4

8.9 9.8 35.4 20.9 2.4 10.9

10 42 5

13 61 6

11 61 4

9 50 2.7

104 528 44

41 190 24.8

14 98 10

45 853 138

86 1294 196

37 1288 172

27 237 38

54 473 74

38 1026 177

84 1313 223

K580 3814 609

CED, cumulative (non-renewable) energy demand; co-combustion in hard coal power plant; reciprocating engine, gasied wood in Otto engine; SRF, short rotation forestry. a Without allocation/credit. b Incineration plant red with wood. c Incl. benzeneCbenzo(a)pyrene.

M. Pehnt / Renewable Energy 31 (2006) 5571 Table 4 Selected inventory and impact assessment results of renewable heat systems Product 1 MJth Unit Forest wood heating plant 61 108 3 4.2 8 5 10 62 124 9 6 4 0.03 0.8 25 6 100 16 SRF heating plant 79 290 2 5.5 12 14 23 68 137 8 10 4 12 0.7 45 10 146 22 Straw heating plant 66 93 2 4.3 19 12 73 181 212 27 7 50 0.03 2.8 143 8 265 28 Forest wood central heating 60 178 4 4.1 17 5 19 75 119 36 28 7 0.03 3.8 191 6 108 15 SRF central heating 76 351 3 5.4 21 13 49 81 131 36 32 7 12 3.8 210 10 169 21 Solar thermal collectors 100 1020 97 6.1 13 0.1 44 32 15 1 13 0.19 0.03 0.01 214 6 54 2

61

Resources CED Iron ore Bauxite Emissions in air CO2 CH4 N2O SO2 CO NOx NMHCa Particles/dust HCl NH3 Benzene Benzo(a)-pyrene Impact assessment Global warming Acidication Eutrophication
a

kJ mg mg g mg mg mg mg mg mg mg mg mg mg Ng g mg mg

Incl. benzeneCbenzo (a) pyrene.

and steel consumption for the protective design, and wind power due to iron consumption for the steel tower. It is necessary to note that other environmental impacts associated with materials supply are included and that, moreover, material input directly depends on local conditions (e.g. concrete input for hydropower plants, aluminum for photovoltaics depending on roof or fac ade integration, etc.). For other environmental impacts no clear trend in results for or against renewable energies arises. In fact, the comparison depends on the technology investigated, the fuel inventory of the used energy carrier (biomass), the specic operational context of the equipment (for example, for the case of photovoltaics, solar insolation, full load hours, topographic site, choice of materials for mounting, etc.), and other relevant factors. By its nature, environmental accounting for renewable energy systems can only provide information about typical systems. For example, the acidication gures for electricity generating systems are well below or similar to the future reference mix, with the exception of the biogas system, which is above the reference mix owing to the ammonia emissions of the agricultural system. Apart from straw as a fuel, the heat generating systems are also below or similar to the reference mix. Straw-red heating plants emit more acidifying substances (chlorine and sulphur content, NOx emissions) than short rotation wood, which in turn emits more than forest wood as a result of the fertilizer and cultivation input and the agricultural emissions.

62

M. Pehnt / Renewable Energy 31 (2006) 5571

The pattern for eutrophication is rather different: electricity generating systems excluding biomass are considerably better than the reference mix, but biomass systems are well above the reference mix (exception: systems with co-combustion of forest wood). This is due in particular to the fact that the NOx emissions of small systems are higher, and that the advantages on the acidication side compared with the reference mix, which result from avoiding the SO2 emissions of coal-red power stations, are not apparent when it comes to eutrophication. On balance, there are thus clear advantages under the headings of greenhouse effect and consumption of nite energy resources. In the other impact categories, the ndings reveal no clear trends. Thus, it is not possible to reach an objective decision. If one considers the great importance for energy resource consumption and greenhouse effect and the great specic contribution of the energy system to these environmental impacts, all renewable energy sources demonstrate clear advantages over the conventional variants where these environmental impacts are concerned. 3. Second step: dynamic LCAs of renewable energy systems 3.1. Methodology The analysis of individual technologies must consider the extremely dynamic development. This concerns the development of products and their production processes as well as their technical performance and the development of so-called background systems (Fig. 2).
Static Analysis
Final energy carrier Material production provision 2010 2010 Final energy carrier provision 2030

Dynamic Analysis
Material production 2030

Transports

Transports

...
Renewable energy technology
Status: Best available technology

Renewable energy technology

Status: Best availabletechnology plus extrapolation of future development

Fig. 2. Dynamic LCAs: principle.

M. Pehnt / Renewable Energy 31 (2006) 5571

63

The following renewable energy carriers are presented here as an example of a dynamic LCA: Photovoltaics (p-Si); Forest timber in central heating; Timber from short rotation forestry used in steam turbines. The following dynamic LCA shall be regarded as an estimate of the order of magnitude of possible impact reductions in the time span, not as an exact forecast. The results are illustrated for the impact categories greenhouse gases and acidication only. 3.2. Dynamic parameters of the background system The future-oriented dynamic assessments are represented and interpreted in the following sections. To present the inuence of the time-dependent parameters, parameter changes are applied for the scenario in 2010 consecutively (cumulative). When interpreting the dynamic assessments one should pay attention to the fact that the results are not commutative, i.e. the order in which the parameters are varied has inuence on the reduction effect, because optimizing an already optimized result has a smaller effect than optimizing the default value. Certainly, the nal result in absolute amount is independent of the sequence of reduction steps. For the dynamic LCA, those parameters are extrapolated into the future which are environmentally relevant and at the same time exhibit a signicant timedependency. The assessment of the system is iterated with those input parameters. With this approach, environmental problem areas, which are inevitably connected with renewable energies, can be analytically distinguished from those that are imported into the system by the background system, i.e. supply of energy and materials. The following parameters are varied: Future power plants (electricity mix 2030). The development of power plants according to a sustainability scenario, which was developed for the Environmental Protection Agency, is analyzed [10]. This scenario, dened by a climate reduction goal of80% by the year 2050, is characterized by signicant contributions from renewable energy carriers. An extrapolation of the efciency and emission development from fossil power plants according to [10] is realized alongside the adapted shares of energy carriers. Aluminum. Future development concerns particularly the reduction of electricity demand for the electrolysis by 7% [11,12]. The recycling share of aluminum depends on the type and composition of the product. On the assembly level, 72% of packaging aluminum, 85% of aluminum in building industry, and 87% of aluminum in electrical engineering are recycled in Germany [13]. 85 and 90% are assumed for 2010 and 2030, respectively. Steel. The present German recycling quota for steel is at a level of 43%. This comprises both own scrap in the steel mills and purchase of external scrap. The assembly based recycling quota depends strongly on the type of steel, the input, the worldwide scrap

64

M. Pehnt / Renewable Energy 31 (2006) 5571

market, etc. The quota of 75% is reported for recycling automobiles. In our assessment, the scrap share is assumed to increase from 46 to 75%. Moreover, the electricity mix 2030 is used for the future steel. Further processes. Further processes are varied specic to technology (e.g. biomass cultivation methods, fertilizers production, increased efciency, process losses at silicon wafer production, etc.). 3.3. Example 1: photovoltaics Future development will lead to a further decrease in production environmental impacts based on the already future-oriented assessment of p-Si, e.g. due to advances in module efciency, improved casting methods and a lower Silicon demand via thinner wafer, reduced saw losses, other production methods, etc. [5,14]. The dynamic parameters are summarized in Table 5. The improvement of production methods and the favorable conditions for materials supply and energy form the basis of these parameters. With regard to the greenhouse effect, each of the rst three dynamic parameters constitutes a decrease in about 20%. Although the production of silicon substantially contributes to the greenhouse effect, the smaller wafer thickness only makes a smaller difference. This is also due to the fact that the improvement step is applied to an already optimized system. For the minimization of acidication, the lifetime and module efciency are of greatest importance (Fig. 3). Overall, the development of optimization potential and the improvement of materials and energy supply allow a 50% reduction of the environmental impacts. Together with quantied optimization steps, there is a possibility to further reduce environmental impacts, in particular in recycling wafer and module components [15]. The recycling of silica could not be quantied here due to the lack of reliable data. 3.4. Example 2: steam turbine power plant with timber from short rotation forestry Today, biomass-red steam turbines often show a very poor performance, with electrical efciencies around 1518%. By 2010, we expect that the efciency of new plants will go up to helZ29% (without cogeneration) in accordance with [16]. In the 2030
Table 5 Parameters varied in the dynamic LCA of p-Si photovoltaics 2010 Steel production Aluminum production Electricity production Life time PV system Module efciency Wafer thickness/sawing loss Scrap share 46%, electricity 2010 Scrap share 85% Business as usual electricity mix 2010 25 years 13.4% 300 mm/200 mm 2030 Scrap share 75%, electricity 2030 Scrap share 90%, reduced electricity demand for electrolysis Sustainable Electricity mix 2030 30 years 17.8% 150 mm/150 mm

M. Pehnt / Renewable Energy 31 (2006) 5571

120% Steel, aluminium and electricity mix 2030 100% 30 years life time 17.8% module efficiency reduced wafer/sawing losses

65

2010 = 100 %

80%

40%

20%

0% 2010 2030 (1) 2030 (2) 2030 (3) 2030 (4)

Fig. 3. Dynamic LCA of photovoltaics for selected impact categories.

sensitivity analysis, this will increase only slightly to 31%. Along with the power plant technology improvement, the improvement of the background system is assumed in analogy to the photovoltaics LCA. Improving European fertilizers and implementing possible measures for emissions reduction from the ground due to fertilizers containing nitrogen are extremely important for the agricultural sector (Table 6). The increase in efciency and reduction of emissions and fertilizer production reduce impacts by 25% points to the benet of the greenhouse effect. The rst aspect is also the most important step for acidication emissions reduction, which alone is decreased by
Table 6 Parameters varied in the dynamic LCA of steam turbines with timber from short rotation forestry 2010 Steel production Aluminum production Electricity production Efciency and emissions of steam turbine power plant Optimized manure production Scrap share 46%, electricity 2010 Scrap share 85% Business as usual electricity mix 2010 CO, NOx, NMHC, particles emission reduction by 20%a helZ29% Reduction of energy demand for manure production by 30%, of CO2 and N2O emissions by 60%b Reduction of NH3 emissions from the eld by 60%c 2030 Scrap share 75%, electricity 2030 Scrap share 90%, reduced electricity demand for electrolysis Sustainable Electricity mix 2030

Global warming

Acidifcation

60%

helZ32%a

Technology for application of liquid manure

a b c

Ref. [16]. Ref. [21]. Ref. [7].

66
120%

M. Pehnt / Renewable Energy 31 (2006) 5571

Steel, aluminium and Effciency, electricity mix 2030 emissions 2030

100%

Manure production 2030 Optimized manure application

2010 = 100 %

80% Global warming 2010

60%

40%

20%

0% 2030 (1) 2030 (2) 2030 (3) 2030 (4)

Fig. 4. Dynamic LCA of steam turbines red with timber from short rotation forestry for selected impact categories.

almost one-fth. In addition, optimizing the application technology of liquid manure allows an acidication gain of 10% points. The changes for materials and energy are insignicant. In conclusion, the technology-specic development of optimization potential of these environmental impacts allows the reduction of some 30% (Fig. 4). 3.5. Example 3: central heating with forest timber Similar to electricity production, implementing technical innovations is important for heat delivering biomass technologies. Stricter legal obligations, particularly in the sector of small systems (e.g. through the planned amendment of technical instructions on air quality control) result in greater efforts of manufacturers to reduce emissions of their equipment. Thus, a signicant reduction of environmental effects can be achieved, especially for air pollutants. The limiting value for dust must be reduced to 100 mg/m3 for devices with a combustion capacity below 2.5 MW that are fed with natural timber from the forest. On the other hand, the required costly ue gas lter technology with electrical lters would generate disproportionately high costs instead of cyclone-principle strippers, which are applied to smaller devices. The dynamic parameters are summarized in Table 7. Like in the case of the abovementioned energy technologies, an improvement of conditions for material and energy supply is assumed. In the case of wood timber in wood chips heating, the materials and energy supply have the strongest impact on the greenhouse effect. The development of the efciency and emissions considerably inuences acidication, whereas the changed supply conditions are hardly relevant. In total, the technology-specic development of optimization potential allows a decrease in 20% of these environmental impacts (Fig. 5).

Acidifcation

M. Pehnt / Renewable Energy 31 (2006) 5571 Table 7 Parameters varied in the dynamic LCA of heat production of wood chips boilers with forest wood 2010 Steel production Aluminum production Electricity production Efciency and emissions of the wood chips boiler
a

67

2030 Scrap share 75%, electricity 2030 Scrap share 90%, reduced electricity demand for electrolysis Sustainable Electricity mix 2030

Scrap share 46%, electricity 2010 Scrap share 85% Business as usual electricity mix 2010 CO, NOx, NMHC, particles emission reduction by 20%a hthZ82%

hthZ84%a

Ref. [16].

4. Expanding the system boundary: effects on the consumer The application of renewable energy sources might not only modify the background system, making ceteris paribus assumptions obsolete. Rather, renewable/distributed energy sources might also modify further downstream aspects, such as consumer behavior. This is particularly the case when renewable energy systems are installed at the customers premises, e.g. on the roof or in the basement of a private household. The emissions reduction and resource protection potential of renewable energy systems could then partially be offset by a rebound effect, thus implying that environmental benets achieved by a more benign technology are at least partly compensated, and sometimes overcompensated, by an increase in energy demand. This rebound effect might be due to [17] behavioral changes, e.g. new comfort features. For instance, the switch from single coal or wood stoves to central heating in residential buildings leads to increases in energy consumption because users increase the number of heated rooms as well as the average
120% Steel, aluminium and electricity mix 2030 100% Efficiency and emissions 2030

2010 = 100 %

80%

Global warning 2010

60%

40%

20%

0% 2030 (1) 2030 (2)

Fig. 5. Dynamic LCA of wood chips boilers with forest wood for selected impact categories.

Acidification

68
2500 2000 1500 1000 500 0 500

M. Pehnt / Renewable Energy 31 (2006) 5571

Electricity demand change (kWh/a)

1000 0 1000 2000 3000 4000 5000 6000 7000 8000

Annual demand (kWh/a)

Fig. 6. Inuence of PV installation on the change in household electricity demand depending on the annual household electricity consumption [19].

temperature. This level of behavioral change depends, among other things, on the relevance of the users ecological norms, behavioral consciousness, the degree to which renewable energy system possession is perceived as ecologically relevant, and knowledge of its effects; increased expenditure available due to saved energy costs; this aspect is generally not relevant in the case of renewable energy systems; off-setting certain symbolic types of environmental action against behavior in other areas (the attitude of now I can drive a car because I have a PV system). On the contrary, installing renewable energy systems could also lead to a stimulated environmental consciousness and enhanced involvement in energy topics. This effect greatly depends on the specic form, timing and detail of feedback, and on the presence of other incentives, such as price incentives, importance of independence, and ecological motives. Whether the rebound effect or the positive effects on environmental consciousness prevails is, however, difcult to quantify and strongly context-dependent. For example, in the case of photovoltaics, Genennig and Hoffmann [18] and Haas et al. [19] have found that electricity consumption rises in households with low prior consumption and decreases in households with high prior consumption (Fig. 6). Apparently, the free energy is used to raise the comfort level of users who were previously deprived of such comfort. In contrast, Haas et al. [20] nd no difference in electricity consumption between households using renewable energies and conventional households. A time perspective on changes in consumption, however, is lacking here.

5. Conclusions From the LCA results it follows that for all renewable energy chains the inputs of nite energy resources and emissions of greenhouse gases are extremely low compared with

M. Pehnt / Renewable Energy 31 (2006) 5571

69

the conventional system. The relevant environmental impacts of the renewable energy systems amount to a maximum of 20% of an expected future German mix for electricity, a maximum of 15% of the reference mix for heat, and a maximum of 55% of the future diesel car in the case of fuels. LCA results for renewable energy systems reveals that the use made of the material resources investigated (iron ore, bauxite) is less than or similar to that made by conventional systems with some exceptions. It should be noted that the other environmental impacts associated with the provision of the materials are of course taken into account, and that the input of materials in particular depends heavily on the local situation. The ndings do not reveal any clear verdict for or against renewable energies in the case of other environmental impacts. The comparison depends more on a large number of context-dependent parameters, e.g. the technology conguration examined (e.g. polycrystalline, monocrystalline or amorphous silicon or thin-lm solar cells, steam turbine or combustion engine CHP units, etc.); the type of energy source used, especially in the case of biomass, and its specic properties (fuel inventory, transport distances, etc.); the geographical location, topographical situation and local conditions of the plant (crucial for solar radiation, full-load hours, expenditure on barrages for hydropower, etc.) and integration into the local infrastructure. Future development will enable a further reduction of environmental impacts that are caused by regenerative energy systems. Different factors are responsible: Progress with respect to technical parameters of the energy converters, in particular improved efciency, emissions characteristics, increased lifetime, etc. Advances with regard to the production process of the energy converters or fuels, e.g. reduced sawing losses or wafer thickness for solar cells, decreased fertilizer input, and higher yields for biomass cultivation, etc. Advances with regard to external services originating from conventional energy and transport systems, for instance improved electricity or process heat supply for system production, ecologically optimized transport systems for the biomass transportation, etc. On the other hand, the last aspect could potentially lead to higher ecological impacts, because the attainable credits for by-products (avoided burden), e.g. glycerin in bio diesel production, are also lower. Nevertheless, the combined effect of the three progress (advance) factors will allow substantial reduction of environmental impacts.

Acknowledgements The work described in the paper was funded under the socio-ecological research framework recently launched by the German Ministry for Education and Research r Bildung und Forschung, BMBF) and applied within a project (Bundesministerium fu

70

M. Pehnt / Renewable Energy 31 (2006) 5571

funded by the German Federal Ministry for the Environment. The author acknowledges rtner (IFEU) for biomass data. support by Sven Ga

References
[1] Pehnt M. Assessing future energy and transport systems: the case of fuel cells. Part 1: methodological aspects. Int J LCA 2003;6:2839. Pehnt M. Assessing future energy and transport systems: the case of fuel cells. Part 2: environmental performance. Int J LCA 2003;6:36578. [2] Hartmann D. Ganzheitliche Bilanzierung der Stromerzeugung aus regenerativen Energien. PhD Thesis. t Stuttgart; 2001. Stuttgart: Universita kologisch optimierter [3] Viebahn P. LCA of solar thermal power plants. Calculations within the project O r Luft- und Ausbau der Nutzung erneuerbarer Energien in Deutschland. Stuttgart: Deutsches Zentrum fu Raumfahrt; 2003. kologische Analyse im Kontext nftige Entwicklung von Stromerzeugungstechnikeneine o [4] Nill M. Die zuku konomischer Zusammenha nge. PhD Thesis. Hamburg: Technische Universita t Hamburg, technischer und o Harburg; 2003. kobilanz Photovoltaikstrom und Update der O kobilanz fu r das [5] Jungbluth N, Frischknecht R. Literaturstudie O Jahr 2000. Uster (Schweiz): ESU Services; 2000. rmeversorgungen. tzten Nahwa [6] Raab S. Ganzheitliche Energie- und Emissionsbilanzierung von solar unterstu t Stuttgart; 2001. Stuttgart: Universita kobilanz der Stromgewinnung aus landwirtschaftlichem [7] Edelmann W, Schleiss K, Engeli H, Baier U. O denswil; r Energie. Bern: arbi GmbH, Engeli engineering, Hochschule Wa Biogas. Report for the Bundesamt fu 2001. kologischer optimierter Ausbau der Nutzung erneuerbarer Energien in [8] DLR, IFEU, Wuppertal-Institut. O Deutschland. Report for the German Ministry of the Environment. Berlin: DLR, IFEU, Wuppertal-Institut. Download from http://www.erneuerbare-energien.de/1024/index.php?fbZ/sachthemen/ee/eu/links/&nZ 11923; 2004. [9] Enquete. Abschlussbericht der Enquete-Kommission Nachhaltige Energieversorgung unter den Bedingungen der Globalisierung und der Liberalisierung. Berlin: Enquete-Kommission; 2002 http://www.bundestag.de/ gremien/ener/index.html. r eine nachhaltige Energieversorgung, Report for the [10] Fischedick M, Nitsch J. Langfristszenarien fu Umweltbundesamt. Forschungsbericht 200 97 104. Wuppertal, Stuttgart; 2002. [11] Rombach G, Zapp P, Kuckshinrichs W, Friedrich B. Technical progress in the aluminium industrya scenario lich; 2001. approach. In: Forschungszentrum Ju [12] EAA. Environmental Prole Report for the European Aluminium Industry. European Aluminium Association; 2000. [13] GDA, Aluminiumindustrie Gd, Recycling. www.aluinfo.de; 2002. [14] Alsema EA. Energy pay-back time and CO2 emissions of PV systems. Prog Photovolt Res Appl 2000;8:1725. [15] Frisson L, Hofkens H, Clerqk Kd, Nijs J, Geeroms A. Cost effective recycling of PV modules and the impact on the environment, life cycle, energy payback time and cost. In: Proceeding of the second world conference on photovoltaic energy conversion. Wien 1998. n D, Reinhardt G, Baur F, Flake M, et al. Stoffstromanalyse zur nachhaltigen [16] Fritsche U, Heinz A, Thra energetischen Nutzung von Biomasse. Project funded by the Federal Ministry of the Environment. Darmstadt: ko-Institut, FhG Oberhausen, IFEU Heidelberg, IZES Saarbru cken, TU Braunschweig, TU Mu nchen; 2003. O [17] Fischer C. Consumer behaviour and consumption patterns. In: Pehnt M, Schneider L, editors. Microcogeneration. Heidelberg: Springer; 2005. nder-1000 Da cher [18] Genennig B, Hoffmann VU. Sozialwissenschaftliche Begleituntersuchung zum Bund-La Photovoltaik-Programm. Umweltinstitut Leipzig: Leipzig; 1996. bner M. Socio-economic aspects of the Austrian [19] Haas R, Ornetzeder M, Hametner K, Wroblewski A, Hu 200 kWp-photovoltaic-rooftop programme. Sol Energy 1999;66:18399.

M. Pehnt / Renewable Energy 31 (2006) 5571

71

ger und Energiever[20] Haas R, Biermayr P, Baumann B, Schrie E, Skopetz H. Erneuerbare Energietra r Verkehr, Innovation und Technologie; 2001. brauchsverhalten. Wien: Bundesministerium fu [21] Kongshaug G. Energy consumption and greenhouse gas emissions in fertilizer production. In: 1998 Technical Conference, Marrakech, Morocco; 1998.

Dr Martin Pehnt was born 8.6.1970 in Cologne and has two children. He studied Physics and Energy Management bingen, Stuttgart, Berlin and Boulder (USA), and obtained a Masters Degree in 1996 and a PhD in Energy in Tu Technology in 2002 (very good). He undertook a research period at the National Renewable Energy Laboratory NREL (Golden, Colorado). From 1997 to 2001 he was a Scientist at the German Aerospace Center, Institute for Technical Thermodynamics, Department for Systems Analysis and Technology Assessment. Since 2001, Mr Pehnt r Energie- und Umweltforschung Heidelberg. His research foci has been a Senior Scientist at the IFEU Institut fu include: Technology and environmental assessment of innovative energy and transport systems, such as fuel cells, cogeneration, renewable energies, biofuels; environmental communication.