Davide Tonini PhD Thesis FULL Version for ORBIT.pdf

8/12/2019 Davide Tonini PhD Thesis FULL Version for ORBIT.pdf

http://slidepdf.com/reader/full/davide-tonini-phd-thesis-full-version-for-orbitpdf 1/236

PhD ThesisFebruary 2013

Davide Tonini

Environmental assessment of energy productionfrom waste and biomass



Environmental assessment of energyproduction from waste and biomass

Davide Tonini

PhD ThesisFebruary 2013

DTU EnvironmentDepartment of Environmental Engineering

Technical University of Denmark



Davide Tonini

Environmental assessment of energy production from waste and biomass

PhD Thesis, February 2013

The synopsis part of this thesis is available as a pdf-file for download from theDTU research database ORBIT: http://www.orbit.dtu.dk

DTU Environment

February 2013

Department of Environmental EngineeringTechnical University of Denmark Miljoevej, building 1132800 Kgs. LyngbyDenmark

+45 4525 1600+45 4593 2850

http://www.env.dtu.dk [email protected]

Vester Kopi

Torben Dolin

Address:

Phone reception:Fax:

Homepage:E-mail:

Printed by:

Cover:



i

PrefaceThis thesis comprises the research carried out for a PhD at DTU Environment,

Technical University of Denmark, from 2009 to 2012. The thesis was funded bythe projects PSO-7335 REnescience, CEESA (Coherent Energy andEnvironmental System Analysis), EUDP 304701 and by Technical University ofDenmark (DTU). The study included collaborations with PhD candidate LorieHamelin (University of Southern Denmark), PhD Gianluca Dorini (TechnicalUniversity of Denmark) and PhD candidate Cristina Montejo (Universidad deSalamanca, Spain). The supervisor was Professor Thomas Astrup.

The PhD thesis comprises a synopsis of the work presented in three published

papers, two submitted papers and one manuscript to be submitted. In the synopsisof the thesis the papers are referred to by the names of the authors and the RomannumeralsI-VI (e.g. Tonini et al.,III ). The papers included in the thesis are:

I. Tonini, D., Astrup, T., 2012. Life-cycle assessment of biomass-basedenergy systems: A case study for Denmark. Appl. Energy 99, 234-246.

II. Tonini, D., Hamelin, L., Wenzel, H., Astrup, T., 2012. BioenergyProduction from Perennial Energy Crops: a Consequential LCA of 12Bioenergy Scenarios including Land Use Changes. Environ. Sci. Technol.46(24), 13521-13530.

III. Tonini, D., Astrup, T., 2012. Life-cycle assessment of a waste refinery process for enzymatic treatment of municipal solid waste. Waste Manage.32, 165-176.

IV. Tonini, D., Dorini, G., Astrup, T. Advanced material, substance and

energy flow analysis of a waste refinery process. Submitted toBioresource Technol.

V. Montejo, C., Tonini, D., Marquez, C.M., Astrup, T. Mechanical-biologicaltreatment: performance and potentials. A LCA of 8 MBT plants includingwaste characterization. Submitted to J. Environ. Manage.

VI. Tonini, D., Martinez, V., Astrup, T. Potential for waste refineries inEurope. To be submitted to Environ. Sci. Technol.



ii

In addition, the following publications have been produced during the PhD:

Clavreul, J., Guyonnet, D., Tonini, D., Christensen, T.H. Comparison of

uncertainty propagation with probability and possibility theories in LCA.Submitted to Int. J. Life Cycle Assess.

Mathiesen, B.V., Lund, H., Hvelplund, F.K., Connolly, D., Bentsen, N.S.,Tonini, D., Morthorst, P.E., Wenzel, H., Astrup, T., Meyer, N.I., Münster, M.,Østergaard, P.A., Bak-Jensen, B., Nielsen, M.P., Schaltz, E., Pillai, J.R.,Hamelin, L., Felby, C., Heussen, K., Karnøe, P., Munksgaard, J., Pade, L.,Andersen, F.M., Hansen, K., 2011. CEESA 100% Renewable Energy Scenariostowards 2050. Aalborg University, Aalborg, Denmark. Available at:http://www.ceesa.plan.aau.dk/digitalAssets/32/32603_ceesa_final_report_samlet _02112011.pdf.

Manfredi, S., Tonini, D., Christensen, T.H., 2011. Environmental assessment ofdifferent management options for individual waste fractions by means of life-cycle assessment modelling. Resour. Conserv. Recy. 55, 995-1004.

Christensen, T.H., Simion, F., Tonini, D., Møller, J. 2011. LCA Modeling ofWaste Management Scenarios. In Christensen TH. (Ed), Solid Waste Technologyand Management. Willey & Sons, London, 161-179.

Manfredi, S., Tonini, D., Christensen, T.H., 2010. Contribution of individualwaste fractions to the environmental impacts from landfilling of municipal solidwaste. Waste Manage. 30, 433-440.

Christensen, T.H., Simion, F., Tonini, D., Møller, J. 2009. Global warmingfactors modelled for 40 generic municipal waste management scenarios. Waste

Manage & Res. 27, 9, 871-884.

Manfredi, M., Tonini, D., Christensen, T.H. 2009. Landfilling of waste:accounting of greenhouse gases and global warming contributions. WasteManage Res. 27, 9, 825-836.



iii

AcknowledgementsFirst, I would like to thank my supervisor Thomas Astrup for convincing me to

apply for the PhD at the very beginning of this adventure, and for latersupporting me during the following years.Warm thanks to Lorie Hamelin for the very good and constructive collaborationswe had, for the constant exchange of information and knowledge, and for her positive and always helpful attitude.A great thank to Thomas Christensen and Simone Manfredi who helped meduring my master thesis and introduced me to the PhD.Special thanks to Henrik Wenzel, Cristina Montejo, Gianluca Dorini, VeronicaMartinez, Julie Clavreul and Maria C. Marquez for their contribution to this PhD

through important scientific collaborations.A great thank to Alessio for his help and support whenever he was asked for it.A very big hug to all my friends and to the closest people I have here: Nemanja,Fabrizio, Roberto, Elena, Manos, Carloto and, of course, Carolina, for her greatand constant smile.Great thanks to Laura and Line for helping translating the abstract. To Torbenand Lisbet for the endless patience in graphics supporting during the years.A grateful thank to my parents for their constant, warm, rich-in-food welcomingevery time I visit them in my hometown.A special thank to my younger sister Debora, for editing some of the figures inthis thesis and for providing me with a good racing bike, which is important inDK. Also, a big hug to my other sister Barbara who is in Australia.Greetings also to my friends in Italy: Bruno, Fabio, Giovanni, Eleonora and allthe others.



iv



v

AbstractOptimal utilization of biomass and waste for energy purposes offers great

potentials for reducing fossil fuel dependency and resource consumption. Thecommon understanding is that bioenergy decreases greenhouse gas (GHG)emissions as the carbon released during energy conversion has previously beencaptured during growth of the plants. This, however, neglects that using the landfor energy crops implies that the same land cannot be used for other purposes,including food cropland, forestry, grassland, etc. This may induce cascadingeffects converting natural biomes into arable land with associated impacts.Waste, such as municipal solid waste, does not involve land use change impacts.However, existing and emerging waste treatment technologies offer different

environmental benefits and drawbacks which should be evaluated in order torecommend appropriate technologies in selected scenarios.

To evaluate the environmental and energy performance of bioenergy and waste-to-energy systems life cycle assessment was used in this thesis. This wassupported by other tools such as material, substance, energy flow analysis andenergy system analysis. The primary objective of this research was to provide aconsistent framework for the environmental assessment of innovative bioenergyand waste-to-energy systems including the integration of LCA with other tools(mentioned earlier). The focus was on the following aspects:

Evaluation of potential future energy scenarios for Denmark. This was done by integrating the results of energy system analysis into life cycle assessmentscenarios.

Identification of the criticalities of bioenergy systems, particularly in relationto land use changes.

Identification of potentials and criticalities associated with innovative wasterefinery technologies. This was done by assessing a specific pilot-plantoperated in Copenhagen, Denmark. The waste refining treatment wascompared with a number of different state-of-the-art technologies such asincineration, mechanical-biological treatment and landfilling in bioreactor.

The results highlighted that production of liquid and solid biofuels from energycrops should be limited when inducing indirect land use changes (iLUC). Solid biofuels for use in combined heat and power plants may perform better thanliquid biofuels due to higher energy conversion efficiencies. The iLUC impacts

stood out as the most important contributor to the induced GHG emissions within



vi

bioenergy systems. Although quantification of these impacts is associated withhigh uncertainty, an increasing number of studies are documenting thesignificance of the iLUC impacts in the bioenergy life cycle.

With respect to municipal solid waste, state of the art incineration, MBT andwaste refining (with associated energy and material recovery processes) may all provide important and comparable GHG emission savings. The wastecomposition (e.g. amount of organic and paper) and properties (e.g. LHV, watercontent) play a crucial role in affecting the final ranking. When assessing theenvironmental performance of the waste refinery, a detailed knowledge of thewaste composition is recommendable as this determines the energy outputs andthereby the assessment results.

The benefits offered by the waste refinery compared with incinerators and MBT plants are primarily related to the optimized electricity and phosphorousrecovery. However, recovery of nutrients and phosphorous might come at theexpenses of increased N-eutrophication and emissions of hazardous substances tosoil. The first could be significantly mitigated by post-treating the digestate leftfrom bioliquid digestion (e.g. composting). Compared with waste refiningtreatment, efficient source-segregation of the organic waste with subsequent biological processing may decrease digestate/compost contamination and recover phosphorous similarly to the waste refinery process. However, recent studieshighlighted how this strategy often fails leading to high mass/energy/nutrientslosses as well as to contamination of the segregated organic waste with unwantedimpurities.

All in all, more insight should be gained into the magnitude of iLUC impactsassociated with energy crops. Their quantification is the key factor determining a beneficial or detrimental GHG performance of bioenergy systems based on

energy crops. If energy crops are introduced, combined heat and power production should be prioritized based on the results of this research. Productionof liquid biofuels for transport should be limited as the overall energy conversionefficiency is significantly lower thereby leading to decreased GHG performances.On this basis, recovery of energy, materials and resources from waste such asresidual agricultural/forestry biomass and municipal/commercial/industrial wasteshould be seen as the way ahead. Highly-efficient combustion and incinerationoffer robust energy and environmental performances. Innovative waste refineriesmay achieve similar performances from a GHG perspective and, in addition, mayrecover nutrients.



vii

In the perspective of future energy systems with increased shares of fluctuatingenergy sources (e.g. wind energy) the flexibility of the energy conversion processshould also be considered in the environmental assessment. The storability of the

produced energy carrier along with the regulation ability and the capacity ofswitching among outputs may offer substantial benefits to the surroundingenergy system. In this perspective, waste refineries producing storable biogas andsolid fuel may offer increased flexibility compared with base load incinerators.



viii



ix

Dansk sammenfatningOptimal udnyttelse af biomasse og affald til energiformål, giver stort potentiale

for en reduktion i afhængigheden af fossile brændstoffer og ressourceforbrug.Ideen er at bioenergi nedsætter udledningen af drivhusgasser, da kulstoffrigivet/udledt gennem energikonverteringen stammer fra biogent kulstoftidligere optaget gennem planters vækst. Dette forsømmer, ikke desto mindre, at brugen af land/jord til energiafgrøder indebærer at den samme jord ikke kan bruges til andre formål som f.eks. landbrugsafgrøder, skovbrug, græsarealer, ect..Det kan skabe en akkumulerende effekt og medfølgende miljøpåvirkninger atomdanne naturlige biomer til agerjord. Affald, som f.eks. kommunalt brandbartaffald, indebærer ikke konsekvenser for ændringen af brug af landbrugs jord.

Dog tilbyder eksisterende og fremspirende/kommendeaffaldsbehandlingsteknologier forskellige energi og miljømæssige fordele ogulemper, som bør evalueres for at kunne anbefale den bedst mulige/optimaleteknologi for det enkelte scenarie.

For at evaluere miljø- og energimæssig ydeevne/præstation af bioenergi ogaffald-til-energi systemer, er der i denne afhandling gjort brug aflivscyklusvurdering (LCA). Dette er understøttet af andre metoder/værktøjer, såsom materiale, stof og energi flow analyse samt energisystem analyse. Det primære formål med denne analyse var at skabe en sammenhængende ramme formiljøvurderingen af innovative bioenergi og affald-til-energi-systemer, herunderintegrationen af LCA sammen med andre, tidligere nævnte, metoder/ værktøjer.Fokus er på følgende aspekter:

Evaluering af potentielle fremtidige energiscenarier for Danmark. Dette blevudført, ved at integrere resultaterne fra energisystemanalysen ilivscyklusvurderings scenarier.

Identificering af svaghederne ved bioenergi systemerne, specifikt i forhold tilændringer i jordbrug.

Identifikation af potentialer samt svagheder forbundet med innovativeaffaldsraffineringsteknologier. Dette blev gjort ved at vurdere et specifikt pilot anlæg, der drives i København, Danmark.Affaldsraffineringsbehandlingen blev sammenlignet med en række forskellige”state of the art” teknologier, så som forbrænding, mekanisk-biologisk behandling samt deponering i bioreaktor.



x

De udvalgte resultater understregede at dyrkningen og produktionen af fast ogflydende biobrændsel fra energiafgrøder (selv flerårige afgrøder) leder tilindirekte ændringer i jordbrug/areal anvendelse (iLUC).

Solide biobrændstoffer til brug på kombinationerede varme ogelektricitetsværker, har muligvis en større effekt end flydende, takket være enstørre energikonverteringseffekt. Disse gjorde sig klart bemærket som devigtigste faktorer til de inducerede drivhusgasudledninger indenfor bioenergisystemer. Selvom deres kvantificering er forbundet med storusikkerhed, viser et stigende antal undersøgelser betydningen af iLUC belastninger på bioenergilivscyklussen..

Med hensyn til kommunalt brandbart affald, kan (state of the art) forbrænding,MBT og affaldsraffinering (med supplerende energi og nyttiggørelsesprocesser)være vigtige og bidrage til sammenlignelige besparelse for drivhusgasudledning.Affaldets sammensætning, f.eks. mængden af organisk stof og papir samtegenskaber, som LHV eller vandindhold spiller en afgørende rolle i den endeligeranking. Ved vurderingen af den miljømæssige præstation afaffaldsraffineringen, er det anbefalelsesværdigt at have detaljeret kendskab tilaffaldets sammensætning, da dette er bestemmende for energiproduktionen ogdermed resultaterne af evalueringen.

Fordelene der tilbydes ved affaldsraffinering sammenlignet med forbrændingsanlæg og MBT anlæg ligger primært på den optimerede elektricitet ogforforgenvinding. Genvinding af næringsstoffer og fosfor kan imidlertid ske på bekostning af øget N-eutrofiering og metal belastning af jorden. Dette kanvæsentlig afbødes ved en efterbehandlingsproces af biomassen produceret af biobrændselsspaltning (f.eks. kompostering).

Sammenlignet med affaldsraffineringsbehandling, kan effektiv kildesortering af

det organiske affald med efterfølgende biologisk behandling nedsætte metalkontaminering samt sikre fosforgenvindingen lig en affaldsraffineringsprocess. Nye studier viser dog hvordan denne strategi ofte fejler og leder til stortmasse/energi/næringsstof tab, såvel som kontaminering af det organiske affaldmed uønskede urenheder.

Alt I alt, større indsigt i betydningen af iLUC påvirkninger i forbindelse medenergiafgrøder, bør efterstræbes. Deres kvantificering er nøglefaktorer i bestemmelsen af positiv og skadelig GHG virkning af bioenergisystemer. Hvisder skal gøres brug af disse, baseret på resultaterne af denne forskning, bør



xi

kombineret varme og kraft/elektricitet prioriteres. Produktion af flydende biobrændstoffer til transport bør begrænses da den overordnedeenergikonvertering er betydelig lavere og indvirkende til en nedsat GHG

præstation. I dette lys, skal genindvinding af energi, materialer og ressourcer fraaffald fra, bolig, landbrug, og skovbrug, og kommunalt, kommercielt, ogindustrielt affald ses som vejen frem. Høj effektiv forbrænding tilbyder en robustenergi og miljøsikker præstation. Innovative affaldsraffinaderier kan opnåsammenlignelige præstationer fra et GHG perspektiv og ydermere kan derudvindes næringsstoffer.

Med henblik på fremtidige energisystemer, med øget behov for varierendeenergikilder f.eks. vind energi, bør fleksibiliteten afenergikonverteringsprocessen også tages i betragtning i forhold tilmiljøvurderinger. Opbevaringsmulighederne af den producerede energi bærersammen med reguleringsmulighederne, evnen og kapaciteten til at skifte mellemoutputs og kan dermed tilbyde forskellige fordele for energisystemet. Set fradette perspektiv, kan affaldsraffinering der producerer biogas og fast brændstofmed muligheder for opbevaring, tilbyde en større fleksibilitet sammenlignet med(base-load) forbrændingsanlæg.



xii



xiii

List of contentsPreface .................................................................................................................... i

Acknowledgements ..............................................................................................iii

Abstract ................................................................................................................. v

Dansk sammenfatning ......................................................................................... ix

List of contents ...................................................................................................xiii

1. Introduction ...................................................................................................... 1 1.1 Definitions .................................................................................................. 1 1.2 Available biomass resource potential ........................................................ 3

1.3 Assessment of bioenergy and WtE systems .............................................. 3 1.4 Objectives of the thesis .............................................................................. 4 1.5 Content of the thesis ................................................................................... 6

2. Method ............................................................................................................... 7 2.1 Life cycle assessment ............................................................................ 8 2.2 Material, substance and energy flow analysis....................................... 8 2.3 Energy system analysis ......................................................................... 9 2.4 Waste sampling and characterization .................................................... 9

3. Key factors in LCA of bioenergy and WtE .................................................. 11 3.1 Critical assumptions ................................................................................. 11 3.2 Key aspects in goal and scope definition ................................................. 15

3.2.1 Functional unit ................................................................................. 15 3.2.2 Temporal, geographical and technological scope ........................... 15 3.2.3 Identification of the marginals......................................................... 16

3.3 Land use changes ..................................................................................... 18 3.4 Sensitivity and uncertainty analysis ......................................................... 20

4. Processes in bioenergy and WtE system - Inventory data .......................... 23 4.1 Agricultural processes .............................................................................. 23

4.2 Storage processes ..................................................................................... 23 4.3 Pre-treatment processes ........................................................................... 25

4.3.1 Pre-treatments for biomass .............................................................. 25 4.3.2 Pre-treatments for MSW .................................................................. 25

4.4 Energy conversion technologies .............................................................. 29 4.4.1 Anaerobic digestion ......................................................................... 29 4.4.2 Pyrolisis and gasification ................................................................. 30 4.4.3 Direct combustion and co-firing of biomass ................................... 31 4.4.4 Incineration and co-firing of MSW ................................................. 32 4.4.5 Liquid biofuels ................................................................................. 33



xiv

5. Environmental performance of bioenergy and WtE systems .................... 35 5.1 Case study: future energy scenarios for DK ....................................... 35

5.1.1 Modeling aspects ............................................................................. 35 5.1.2 Key results ....................................................................................... 37

5.2 Case study: bioenergy from perennial crops ....................................... 38 5.2.1 Modeling aspects ............................................................................. 38 5.2.2 Key results ....................................................................................... 39

5.3 Case study: energy from MSW ........................................................... 42 5.3.1 Modeling aspects ............................................................................. 42 5.3.2 Key results – focus on the waste refinery ........................................ 43 5.3.3 Key results – focus on MBT ............................................................ 47

6. Discussion ........................................................................................................ 49

7. Conclusion and recommendations ................................................................ 55

8. Perspectives ..................................................................................................... 57

9. References ........................................................................................................ 59

10. Papers ........................................................................................................... 69



1

1. IntroductionRecovery of energy, materials and resources from different waste and biomasstypes offers great potentials to reduce resources depletion, fossil fuelsconsumption and related environmental burdens. However, determining theoptimal environmental strategy for waste and biomass use for energy purposes isnot straightforward. A number of factors may induce environmental impacts, e.g.shifting the burden from one environmental compartment to another. Theintegration of life cycle assessment with material, substance and energy flowanalysis provides a comprehensive and holistic basis to evaluate theenvironmental performance. The focus of this thesis is on the environmentalassessment of bioenergy and waste-to-energy systems. Special attention is

devoted to emerging solid and liquid biofuels and innovative waste treatmenttechnologies, e.g. waste refineries.

1.1 DefinitionsBelow follows a list of the relevant terminology used within this thesis. Therelative definition as intended within this thesis is given.

Biomass : the biodegradable fraction of products, waste and residues from

agriculture (including vegetal and animal substances), forestry and relatedindustries, as well as the biodegradable fraction of industrial and municipal waste(The European Parliament and The Council, 2008). Within this thesis it is used toindicate both residual agricultural biomasses and energy crops when a distinctionis not needed.

Bioenergy : energy produced from biomass.

MSW : municipal solid waste, i.e. waste from household, as well as other waste,which, because of its nature or composition, is similar to waste from households

(CEC, 1999). Process (or unit process ): a step in manufacturing where a transformation(chemical, physical) takes place (Austen, 1984). For example, cultivation,transportation, pre-treatment, refining, incineration is a process. Each individual process could also comprise a number of sub-processes. The use of the term inthis thesis depends on the context: for example the waste refinery is a process within the waste management chain. The enzymatic treatment is a process withinthe waste refinery.



2

Residual (agricultural) biomass : biomass that is not specifically produced forthe market (i.e. not prime product of an activity). For example straw left-overfrom harvests, grass from uncultivated fields or low lying areas, animal manure,

wood residues, forest residues, etc. Note that the term residual does not implythat the biomass has not a current function in the ecosystem or in the society.

Residual MSW : within this thesis it is used to indicate the residual share of MSW (rMSW) left-over after household source-segregation.

Resource : available source of wealth; a new or reserve supply that can be drawnupon when needed (Oxford Dictionaries, 2012). Within this thesis it is often usedin relation to biomass and phosphorous.

Scenario : projection of a system within the life cycle assessment (LCA). Anyspecific system assessed within the LCA is defined as scenario . For instance the projection of the Danish energy system in 2050 is a scenario .

System : an assemblage or combination of parts forming a complex unitary whole(Oxford Dictionaries, 2012). While scenario refers to a specific LCA projectionof a system, system is used in general terms to indicate any combination of wasteand biomass processes forming a chain.

Technology : the application of scientific knowledge for practical purposes,especially in industry. Within this thesis it is often used as synonymous for process to generally indicate engineering applications. For example a wasterefinery, an incineration, a flue-gas cleaning, a mechanical-biological treatment isa technology .

Waste : materials that are not prime products (i.e. products produced for themarket) for which the generator has no further use for own purpose of production, transformation or consumption, and which he discards, or intends oris required to discard. Wastes may be generated during the extraction of rawmaterials during the processing of raw materials to intermediate and final products, during the consumption of final products, and during any other humanactivity. The following are excluded: i) residuals directly recycled or reused atthe place of generation (i.e. establishment); ii) waste materials that are directlydischarged into ambient water or air (OECD, 2012). Within this thesis it is usedto indicate both MSW and residual MSW when a distinguee is not needed.



3

1.2 Available biomass resource potential

In the perspective of decreasing fossil fuel consumption the available biomassresource potential for energy acquires increasing importance. Estimates for EU27 by Panoutsou et al. (2009) suggest that the available biomass resource potentialcorresponded to about 6.7 EJ y-1 in 2010 and could increase up to 7.8 EJ y-1 by2020. In 2010 agricultural residues had the largest share (2.3 EJ y-1, i.e. 34% ofthe total) followed by forestry residues (1.9 EJ y-1, i.e. 29%) and biodegradablewaste (1.3 EJ y-1, i.e. 19%). Animal manure represented about 11% of the total(corresponding to 33% of agricultural residues). For the specific case ofDenmark, Jørgensen et al. (2008) reported a potential of about 142 PJ y-1 in 2008.This excluded biodegradable waste which might represent additional 20 PJ y-1 as

estimated by Panoutsou et al. (2009). Similarly to EU27, the largest contributioncame from forestry products and residues (wood pellets, wood chips and woodresidues, 60 PJ) and agricultural residues, primarily straw and animal manure,which potential was estimated to ca. 34 and 23 PJ y-1, respectively. As mentionedearlier, the potential of biodegradable waste (ca. 20 PJ) is also remarkable.Additionally, if waste materials containing fossil carbon are accounted for, the potential associated with waste might raise to ca. 30 PJ y-1. It should be noted as,for the case of EU27 the available biomass potential only corresponded to ca.10% of the total primary energy supply in 2009. This was ca. 69 EJ excluding

international aviation and marine bunkers, based on IEA (2011). For the case ofDenmark the potential was ca. 20% of the primary energy supply that equalled804 PJ excluding international aviation and marine bunkers, based on IEA,(2011); if those additional consumptions were included, the primary energysupply would raise to ca. 864 PJ (DEA, 2009). These data highlight theconstraints of the available biomass potential in relation to the current Countriesneeds. With respect to waste-to-energy (WtE) in Denmark, in 2009 the energyrecovery accounted for about 4% and 20% of the total Danish electricity and heat production, respectively (DEA, 2010a). The vast majority of the recoveredenergy (98%) was produced by incineration.

1.3 Assessment of bioenergy and WtE systems

For the assessment of bioenergy systems, life cycle assessment (LCA)inescapably appears the most appropriate tool for its holistic perspective (amongthe others: Edwards et al., 2008, Cherubini et al., 2009, Cherubini and Strømman,2011). However, LCA of bioenergy systems presents a number of challenges anduncertainties primarily related to the quantification of land use changes (LUC), in



4

particular indirect land use changes (iLUC), induced by crops cultivation. Theresearch on iLUC is still at an early stage and within the LCA community there isno agreement about the accounting methods. However, a number of recent

studies have highlighted how accounting of iLUC might lead to a net greenhousegas impact of the bioenergy compared with the fossil fuel reference system,therefore changing the perception on biofuels (Searchinger et al., 2008,Searchinger, 2010). It should also be noted that, though reflecting the principlesof life cycle thinking, the GHG accounting method suggested by the EU directiveon bioenergy (European Union, 2009) does not completely represent an LCA- based approach (e.g. the term LCA is not even mentioned).

A number of tools exist for the assessment of WtE systems. As thoroughlyreviewed in Pires et al. (2011) and Finnveden et al. (2007), these include: lifecycle assessment (LCA), material, substance and energy flow analysis (MFA,SFA and EFA), environmental risk assessment (ERA), environmental impactassessment (EIA), strategic environmental assessment, energy system analysis(ESA), exergy analysis, entropy analysis, cost-benefit analysis (CBA), life cyclecosting (LCC) and multi-criteria analysis (MCA). The latter integrates a numberof different analyses (e.g. LCA, MFA, ERA, etc.). CBA and LCC focus on socio-economical aspects whereas the remaining on the environmental and energy performance. Finnveden et al. (2007) recommended LCA as most suitable forcomparing environmental performances of alternative waste managementsystems. A life cycle thinking approach is also recommended by The EuropeanParliament and The Council (2008). Further integration of LCA with MFA, SFAand EFA to increase the robustness of the assessment is also an option (Pires etal., 2011, Chen et al., 2012). A number of studies combined LCA and MFA toindividuate optimal management strategies (e.g. Chen et al., 2012, Andersen etal., 2010, Arena et al., 2009).

1.4 Objectives of the thesisThe overall aim of the thesis is to provide a systematic framework for theenvironmental assessment of WtE and bioenergy systems with particular focuson emerging WtE technologies (e.g. waste refineries) and biofuels. This isfinalized at providing scientifically sound recommendations for facilitatingdecision-making processes that involve management of waste and biomass forenergy. The objectives can be summarised as follows:



5

Provide a framework for environmental assessment of bioenergy and WtEsystems including integration of LCA with MFA, SFA, EFA and ESA.

Evaluate potential future energy scenarios for Denmark.

Identify the criticalities of bioenergy systems.

Identify potentials and criticalities associated with innovative waste refinerytechnologies.

Recommend best practices for bioenergy and WtE based on the aboveelements.

The bioenergy research mainly focused on Danish conditions and related casestudies. The focus was placed on the criticalities of future Danish energyscenarios based on a high share of biomass. Further investigations dealt with theenvironmental performance of bioenergy systems based on perennial energycrops. A number of tools were used for the investigations; these included MFA(for mass balances), SFA (for carbon and nitrogen flows), EFA (for energy balances), ESA (for designing scenarios) and LCA.

With respect to WtE special attention was devoted to the waste refinery as anexample of emerging technology optimizing energy recovery. A number ofassessment tools were used to evaluate its environmental and energy performance. These included LCA, MFA, SFA, EFA as well as experimentalwork involving waste sampling and characterization. The waste refinery wascompared with a range of waste treatment processes such as state-of-the-artincineration, mechanical-biological treatment (MBT) and landfilling. Further,specific assessments were performed on a number of existing MBT plants. Thereason for this is that these technologies have undergone a significant proliferation in the last two decades and they, as the waste refinery, represent analternative pre-treatment for residual municipal solid waste (rMSW) prior toenergy recovery and final disposal.



6

1.5 Content of the thesisThe structure of the thesis is as follows:

Chapter 2: Describes the methodological approach used to assess WtE and bioenergy systems. The tools (e.g. LCA, MFA, SFA, EFA, waste samplingand characterization) used to this purpose are also described.

Chapter 3: Discusses key factors in LCA of bioenergy and WtE systems.

Chapter 4: Describes relevant processes and technologies associated with bioenergy and WtE systems.

Chapter 5: Identifies potentials and criticalities associated with bioenergy andWtE systems. Special attention is devoted to the assessment of future energyscenarios, bioenergy systems based on perennials and strategies for thetreatment of MSW based on waste refinery, MBT and incineration.

Chapter 6: Highlights and discusses the most important findings of theresearch based on Chapters 2-5. The chapter elaborates on the findings of theenclosed papers.

Chapter 7: Concludes on the outcomes of the thesis.

Chapter 8: Identifies and discusses issues and topics that could be subject of

further scientific investigations.



7

2. Method

The approach used to perform environmental assessment of bioenergy and WtE

systems was based on a combination and integration of a number of methods: i)LCA, ii) MFA, iii) SFA, iv) EFA (including a variety of energy balances) and v)ESA. Additionally, waste sampling and characterization (vi) was also performedfor a specific case study. The ESA was not actively performed during thisresearch. However, in Tonini and Astrup (I) the output of ESA was used as basisto design the future Danish energy scenarios. In Tonini et al. (II ) MFA, SFA andEFA were extensively used to support the LCA. In Tonini and Astrup (III ) andTonini et al. (VI ) the LCA was performed along with a number of differentenergy balances. In Tonini et al. (IV ) waste sampling and characterization was

combined with MFA, SFA and EFA to illustrate the performance of a wasterefinery process. An overview of the methods utilized in the individual papers is presented in Table 1.

Table 1 . Methods used in the papers that constitute the basis for this thesis.

Paper Study subject matter Methods

I Assessment of future energy scenarios for Denmark withhigh share of biomass LCA, ESA, EFA

II Assessment of bioenergy production from perennialenergy crops in Denmark LCA, MFA, SFA, EFA

III Assessment of a waste refinery process and comparisonwith a Danish incinerator LCA, EFA

IV Material, substance and energy analysis of a wasterefinery process

Waste characterization,MFA, SFA, EFA

V Assessment of MBT-based management strategies inCastilla y Leon (Spain) LCA, MFA

VI Evaluation of the potential benefits associated with wasterefineries in a EU context LCA, EFA



8

2.1 Life cycle assessment Life cycle assessment is a standardized methodology commonly used forevaluating WtE and bioenergy systems. LCA allows for a holistic and systematicassessment of both direct and indirect impacts as well as resources consumption.The LCAs presented in this study were performed using consequential life cycleassessment (ISO, 2006a and ISO, 2006b). The term consequential refers to theaim of the LCA: this should highlight the environmental consequences of adecision (for example deciding between alternative A, B, C, etc.). LCA consistsof four phases: 1) goal and scope definition, 2) inventory analysis, 3) impactassessment and 4) interpretation. The first phase includes specification of the aimof the LCA and definition of system boundaries and functional unit (the unitwhich qualitatively and quantitatively describes the service provided by thesystem under assessment). This phase also includes the specification of thetemporal, geographical and technological scope considered. In the second phase,all relevant direct and indirect emissions associated with upstream anddownstream processes are collected and listed based on the functional unit. In thethird phase the emissions are characterized and aggregated conformingly with theconsidered impact categories. In the fourth the results of the impact assessmentare interpreted and discussed in the perspective of the goal and scope defined inthe first phase.

2.2 Material, substance and energy flow analysisMFA, SFA and EFA are useful techniques to assess mass, energy and substanceflows in a range of different systems (Brunner and Rechberger, 2004, Brunner,2012, Cencic and Rechberger, 2008). In the specific context of wastemanagement MFA, SFA and EFA are often utilized to highlight the fate ofimportant materials and substances and to further suggest system improvementson the basis of the results. EFA is typically used to identify relevant energy flows

within the system under assessment (e.g. energy losses, energy content of wastematerials, energy recovery, etc.). All the MFA, SFA and EFA were facilitatedwith the software STAN (Cencic and Rechberger, 2008). This allowed, amongthe others, to consider the uncertainties inventoried on the most sensitive parameters and to reconcile the data when necessary, based on the proceduredescribed in Cencic and Rechberger (2008).

A particular case of EFA is represented by life cycle energy balances. Anexample is in Tonini and Astrup (III ). Life cycle energy balances account for allenergy-related inputs and outputs (electricity, heat, fuels, including energy



9

required to extract and produce the fuel) to and from the system underassessment. Eventual energy savings associated with avoided production ofenergy from fossil fuels, virgin materials and mineral fertilizers are included. The

results should be expressed as primary energy consumed or saved relative to thefunctional unit of the assessment. As an example, if the net electricity savingassociated with recycling of a selected material equalsε (kWh t-1 material) andthe net heat saving equalsφ (MJth t-1 material), the primary energy saving (MJ t-1 material) would be:

)η1

φβ(φη1)

η3.6

εα(εη3.6savingenergyPrimary

ththelel

Where ηel and ηth are the electricity and heat efficiency of the energy production(MJel MJ-1 fuel and MJth MJ-1 fuel), α and β the electricity and heat consumptionto extract and produce the fuel used for energy production (kWh MJ-1 ‘fuelextracted and produced’ and MJth MJ-1 ‘fuel extracted and produced’). DedicatedLCA softwares might also provide this information.

2.3 Energy system analysisEnergy system analysis (ESA) focuses on design and evaluation of potentialenergy scenarios for a selected region or Country (Lund, 2010, Lund andMathiesen, 2009, Lund, 2007). This is often used in the perspective of increasedshares of biomass and windenergy within the energy system. A range of models performing ESA exist. Among the others, EnergyPLAN is a computer model forhour-by-hour simulations of complete regional or national energy systems,including electricity, individual and district heating, cooling, industry andtransportation (Lund, 2010). ESA was not actively performed within this thesis.However, the output of a specific energy system analysis facilitated withEnergyPLAN was used as basis for LCA in Tonini and Astrup (I).

2.4 Waste sampling and characterizationWaste sampling and characterization was performed within this thesis to improvethe knowledge about the solid and liquid outputs of a specific technology (wasterefinery). The waste sampling on field followed an original procedure due to thespecifity of the technology and of its outputs (see Tonini et al. (IV )). A numberof selected waste material fractions were hand-sorted from the sampled waste.

The chemical characterization of the hand-sorted waste material fractions was



10

performed conformingly with the approach described in Riber et al. (2007) andRiber et al. (2009): the selected waste material fractions were dried and grindedusing appropriate equipment. Further mixing and mass fractional reduction was

performed until the mass required for chemical analysis was obtained. Selectedchemicals were then analyzed; these included fossil carbon content (represented by the14C content in12C) analyzed by accelerated mass spectrometry (AMS).



11

3. Key factors in LCA of bioenergy and WtE3.1 Critical assumptionsLCA of WtE and bioenergy systems typically applies the principles ofconsequential LCA (section 2). Since the aim is evaluating the consequence of adecision inducing a change from the current management of selected biomasses/wastes (reference scenario) to another strategy, the assumptions aboutthe reference scenario become crucial. Today most biomasses have a function inthe ecosystem or in the economy meaning that the utilization of these for energywould induce changes in the ecosystem or in the society if status quo is to bemaintained. As a consequence, the use of available biomass resources for energy purposes instead of the current use (e.g. feeding, bedding, ploughing back tofields etc.) may finally lead to a competition between energy and other uses. Theconsequences of diverting biomass resources to energy production must beaddressed in the LCA. These might include land use changes (as for energycrops), increased fertilizers use, reduced soil carbon stock, etc.

When energy crops are considered, any upstream impact associated withcultivation must be included. The most critical is the quantification of direct andindirect land use changes (dLUC and iLUC). The fundamental assumption is that

using land for energy crops typically implies that this land is not producing plantsfor other purposes, including carbon otherwise sequestered (Edwards et al., 2008,Cherubini et al., 2009, Searchinger et al., 2008, Searchinger, 2010, EEA, 2011).The reference scenario when assessing energy crops should therefore be thecurrent management of the land (e.g. forestry, food crops, etc.). Many previousLCA studies on bioenergy failed in assessing bioenergy systems as they did notinclude iLUC (Searchinger et al., 2008, Searchinger, 2010, EEA, 2011). As anexample, if energy crops replace forest stocks (which would otherwise sequestermore carbon compared with the crops), they may end up increasing theatmospheric carbon concentration. If energy crops displace food crops, this maylead to more hunger if the displaced food crops are not cultivated somewhere else(i.e. using other land previously uncultivated) or, more likely, to emissions fromland use changes if they are. In other words, the reduced fossil carbon emissionthrough fossil fuels replacement might come at the expenses of increased biogenic carbon release from vegetation and soil. This concept is exemplified inFigure 1: this illustrates the consequences of using the land for energy instead offood.



12

If MSW is considered, the reference scenario for Danish conditions should beincineration, this being the reference technology for MSW treatment. Thisimplies that any innovative MSW treatment should be compared with

incineration. However, this is not the case for many other EU regions. Forexample, strategies based on mechanical-biological treatment and/or landfillingare largely practiced in EU. In general, when the geographical scope of the LCAfocuses on specific regions or Countries, the choice of the reference scenarioshould always be based upon local conditions.

When assessing waste management systems the ‘zero burden’ approach istypically applied: all upstream emissions associated with generating the waste areomitted from the LCA (e.g. Clift et al., 2000). This means that any treatment

recovering energy, materials and resources from the waste might determineenvironmental savings compared with ‘not doing anything’. However, asmentioned earlier, ‘not doing anything’ is typically not the reference (at least forEU Countries) as current management practices already exist. Therefore, anyalternative management strategy must be compared with the reference and would

be better only when additional environmental savings are raised. The ‘zero burden’ approach does not apply to bioenergy systems based on energy crops asthese are prime products specifically produced for the market. This implies thatall impacts associated with cultivation and production must be included, asearlier discussed.

Figure 1 . Illustration of the induced consequences of bioenergy production from land

previously dedicated to food crops production (from Tonini et al. ( II )).



13

Figure 2 outlines a possible process flow diagram for the assessment of bioenergy and WtE systems. In grey colour are the contributions which are notessential to perform the LCA (but recommended): ESA, MFA, SFA and EFA

represent complementary tools contributing to increase the overall robustness andtransparency of the environmental assessment. Particularly, MFA, SFA and EFAhighlight relevant materials, substance and energy flows within the system underassessment. This may contribute to identifying key processes where the majorresearch efforts during the inventory phase (2) should be focused on. In addition,the flows quantified with MFA, SFA and EFA may be useful for the resultsdiscussion (3). Uncertainty analysis (see section 3.4) is also recommended;however, performing the uncertainty analysis is often limited by the availabilityand quality of parameters uncertainty data (Clavreul et al., 2012).



14

Figure 2 . Process flow diagram for the assessment of bioenergy and WtE systems.Contributions not essential to perform the LCA (but recommended) are marked grey.Contributions not needed in WtE systems (“zero burden” approach) are marked as dashed boxes.



15

3.2 Key aspects in goal and scope definitionThis phase includes definition of the aim of the assessment, functional unit,temporal, geographical and technological scope, marginal technologies, system boundary and impact assessment method. Table 2 presents an overview of therelevant methodological assumptions in the different papers.

3.2.1 Functional unitAccording to Cherubini et al. (2009), the functional unit (FU) should be the unitof land (e.g. hectare) for LCAs focusing on energy crops as the land representsthe bottleneck; instead, LCAs focusing on waste should be expressed per unitinput (e.g. one tonne) or unit output (e.g. MJ energy). LCAs focusing on

transport biofuels should be expressed per km basis. Tonini et al. (II) usedhectare of land as functional unit to assess bioenergy systems based on energycrops. In Tonini and Astrup (I) the functional unit was instead the provision ofthe primary energy required to satisfy society needs in a number of future energyscenarios. This was needed in order to compare scenarios having different energyinputs and outputs though providing the same service. In the remaining LCA papers focusing on MSW the functional unit was 1 tonne of wet waste.

3.2.2 Temporal, geographical and technological scopeTemporal, technological and geographical scope refers to the dimensions for theuse of the LCA (where and when). Their definition is fundamental as it affectsthe choice of technologies (e.g. efficiencies) and marginals (e.g. energy production). As an example, if the goal of the study is to assess future scenarios,the technology efficiency should be subject to forecasting; if a future energyscenario for a selected region is investigated, primary energy supplies should be projected according to assumptions regarding improved efficiency of buildinginsulation, transportation means, power plants, district heating network as well as

eventual changes in people habits (e.g. diversion of passengers transport to trains, bicycles, etc.). With respect to this, energy system analysis may provide the basisfor further environmental assessments. This approach was followed in Tonini andAstrup (I). An overview of the temporal assumptions within this thesis can befound in Table 2. Notice that the temporal scope of the assessment should not beconfused with the global warming (GW) horizon which simply reflects themethod used to characterize the GW emissions (the gases decays are differentwithin 20, 50 and 100 years). Typically, a 100 year GW time horizon is used.



16

3.2.3 Identification of the marginalsThe marginal technology (or product) is defined as the technology (or product)that is most likely to react to a marginal change in demand or supply of a selected

technology (or a selected product) (Weidema, 2003, Weidema et al., 1999). Inthe consequential LCA approach the system under assessment, whenever producing energy and products, is credited with the avoided emissions associatedwith substitution of the assumed marginal energy and products. In the assessmentof WtE and bioenergy systems the identification of the marginal energytechnology, crop and (in minor extent) fertilizers are of crucial importance as thischoice (particularly for marginal energy and crop) affects the magnitude of theenvironmental savings and/or impacts associated with the system underassessment. A typical approach in LCA of bioenergy and WtE is to assume thatin the long-term energy produced from biomass and waste would lead to thedecommissioning of fossil based energy production capacities (both electricityand heat) as these technologies are generally intended to be phased out in order tocomply with political CO2 reduction targets (e.g. Weidema et al., 1999, Ekvalland Weidema, 2004, Finnveden et al., 2009 among the others).

With respect to the Danish market for electricity, there is a broad consensus inassuming coal as marginal source (Weidema, 2003). Fruergaard (2010)

recommended 2 approaches for identifying the long term marginal: i) based onenergy system analysis; ii) based on policy targets. Based on the first approachMathiesen et al. (2009) suggested coal and wind power as long term marginalsfor Denmark. Based on the second approach, the long term marginal in Denmarkwould be the least environmentally desirable technology, i.e. coal-fired power plants. However, this might not be the case for other EU Countries; for example,Turconi et al. (2011) identified natural gas as marginal for Italy. Within thisthesis coal-based electricity was always assumed as marginal when thegeographical scope was Denmark (e.g. Tonini et al. (II ), Tonini and Astrup

(III )). This assumption was always tested in the sensitivity analysis (or directlyin the baseline) by substituting electricity produced from natural gas-fired power plants. In Montejo et al. (V) the marginal for Spain was natural gas. When thegeographical scope was Europe (Tonini et al. (VI )) coal was assumed asmarginal; the influence of this choice was tested in the sensitivity analysis.

As opposed to electricity, the market for heat is rather local and substitution ofdistrict heating or heating fuels often depends on local conditions and productioncapacities connected to the district heating network in question (Fruergaard et al.,



17

2010a). However, other approaches exist. For example, a market-based analysis by DEA, (2010b) suggests that natural gas should be considered as marginal forDanish conditions. In general, the choice of the marginal for heat is subject to a

high uncertainty especially when the geographical scope of the LCA study is notfocused on a specific region for which the heat market is known. Therefore, theapproach used within this thesis was to assess multiple energy scenariosincluding coal and natural gas directly in the baseline of the LCA or,alternatively, if one fuel (e.g. coal) was assumed as marginal in the baseline, toregularly test this assumption in the sensitivity analysis. This approach allowsassessing the two ends of the range with respect to GHG emissions associatedwith the marginal (fossil) heat source.

With respect to crops cultivation, spring barley is generally considered as themarginal crop for Danish conditions. This is supported by a number of studies(Weidema, 2003, Schmidt, 2008, Dalgaard et al., 2008). A recommendableapproach within the LCA is to test this choice by substituting other crops. Forexample, in Tonini et al. (II ) this choice was tested in the sensitivity analysis bysubstituting winter wheat.

Common practice in consequential LCAs is to consider the digestate producedfrom anaerobic digestion of biomass and organic waste as substitute for mineralfertilizers thus avoiding their production and use. For Danish conditions, recentLCA studies (Hamelin et al., 2011, Hamelin et al., 2012) suggest that calciumammonium nitrate, diammonium phosphate and potassium chloride should beconsidered as marginal N, P and K fertilizers. Other studies (Hansen et al., 2006)suggest instead using average European LCI data.



18

3.3 Land use changesCultivation of energy crops requires use of land thereby inducing direct andindirect land use changes (dLUC and iLUC) under the fact that land available forcultivation is constrained. In the Danish context, the direct land use change(dLUC) consequences reflect the environmental impacts/savings of cultivatingselected energy crops instead of the marginal (e.g. spring barley for Danishconditions). In other Countries, this might come at the expenses of uncultivatedland (pasture, forest, etc.). The environmental impacts/savings associated withthe (avoided) cultivation of the marginal crop must therefore be included in theassessment. This involves the use of inventory data for the cultivation phase aswell as for the variation of soil organic carbon content (∆SOC or SOC changes)

between cultivating the marginal and the substituting crop. This reflects thevariation in the carbon balance (above- and below-ground) of the landconsidered. St-Clair et al. (2008) reports SOC changes associated with thecultivation of different energy crops including short rotation coppice (e.g.willow) and rapeseed. Schmidt (2007) details the SOC associated with rapeseedcultivation in Denmark and in other regions (e.g. Canada). Hamelin et al. (2012)details SOC changes related to the establishment of a number of different annualand perennial crops in Denmark. These data were applied in Tonini and Astrup(I) and Tonini et al. (II ).

The iLUC consequence corresponds to the environmental impact of convertingland nowadays not exploited for crop cultivation to cropland, as a result of theinduced demand for the displaced marginal crop. To quantify this impact, twosteps are needed: i) estimate the amount of land converted and the correspondinggeographical region; and ii) identify the biome types converted. Differentapproaches exist for the quantification of iLUC. A comprehensive overview of partial and general equilibrium models that can be used to model iLUC is givenin Edwards et al. (2008). However, most studies to date focus on biofuelmandates for a variety of shock sizes (e.g. Edwards et al., 2008, Edwards et al.,2010), and as such are difficult to be used directly for other applications. Withinthis thesis three approaches that may find a broader application in LCA studieswere identified.

(1) Schmidt (2008) details a number of possible scenarios associated withincreased biofuel production in Denmark (wheat displacing spring barley).According to the author, the scenario that is the most likely to occur is

conversion of grassland (corresponding to 69% of the land displaced in



19

Denmark) along with intensification of barley cultivation (corresponding to 31%of the land displaced in Denmark) in Canada. This approach was followed inTonini and Astrup (I). The largest uncertainty of this approach is related to the

assumption of market elasticity equal to 1 (i.e. all the Danish barley displaced isreplaced leading to a substitution ratio of 1). This may not be the case as variouseconomical mechanisms may determine a substitution ratio lower than 1. Inaddition, the method used for the choice of the region subject to landdisplacement (Canada) is also not well supported. Schmidt (2008) based thisassumption on the prediction from FAPRI (2006): based on this, Canada would be the region facing the largest increase in barley production in 2005-2016 andwas thus identified as the marginal barley supplier reacting to decreases in theDanish barley supply. The mechanisms leading to land conversion are in realitymore complex and subject to conditions and constraints which typically requirethe support of partial/general equilibrium models.

(2) A different approach was instead used in Tonini et al. (II ): the results ofKloeverpris (2008) for a unitary increase of wheat consumption in Denmark wereused as a proxy to estimate the amount, location and biome types of the landconverted as an effect of the decreased spring barley supply from Denmark.These were obtained by using a modified version of the general equilibriumGTAP model (GTAP, 2012). This implicitly assumes market elasticity lowerthan 1 (substitution ratio < 1, i.e. not all the Danish barley displaced is replaced).Further, the soil and vegetation carbon data from the Woods Hole ResearchCentre (Searchinger et al., 2008) were used to calculate the CO2 emitted based onthe methodology reported in Müller-Wenk and Brandao (2010) (CO2 emissionsassociated with the land conversion were not estimated in Kloeverpris, 2008). Note that the results in Tonini et al. (II ) only covered the iLUC impactsassociated with land conversion; the impacts associated with intensification ofthe current cultivation practices (which Kloeverpris, 2008 indicate as

corresponding to ca. 30% of the response to the initial displacement) were notincluded. This indicates that the actual overall iLUC impact may be higher.Though high uncertainty is inherently associated with the estimates on landconversion due to the complexity of the mechanisms involved, the GTAP reflectsthe entire global economy and it is thus well suited for the analysis of globalconsequences of changes in crop demand (Kloeverpris, 2008).

(3) Fritsche (2008) uses a simplified deterministic approach to estimate averageiLUC impacts. The basic assumptions are that: i) current patterns of land use for producing traded agricultural commodities are an adequate proxy to derive global



20

averages for GHG emissions from iLUC; ii) future patterns for global trade can be derived from observed trends. Many uncertainties are associated with thismethod. For example, the analysis considers only key Countries (e.g. US, EU,

Argentina, Brasil, Indonesia, etc.). In addition, the choice of the biomes affectedrelies on arbitrary and not well supported assumptions. However, the relatedresults were used in this thesis for the purpose of comparison (see section 6).

3.4 Sensitivity and uncertainty analysisThe uncertainty of assumptions (e.g. marginals) and parameters (e.g. crops yield,LHV, efficiencies, etc.) used in the LCA requires to be tested in sensitivity anduncertainty analysis. Within this thesis, the term sensitivity analysis has been

used to indicate testing scenario uncertainties, whereas uncertainty analysis toindicate assessing parameters uncertainty. This distinguee is adapted fromHuijbregts et al. (2003) where the authors classify uncertainties in LCA as: i)model uncertainties, ii) scenario uncertainties and iii) parameter uncertainties.The first (i) is associated with the models and equations used to quantify theemissions flows and with the impact assessment methodology selected which provides the characterization factors for relating the inventoried emissions toenvironmental impacts. Scenario uncertainties (ii) are related to uncertaintiesassociated with the choice of technologies and processes and to the fundamentalassumptions intrinsically connected to the consequential LCA approach, i.e. theassumptions for the marginals. Finally, parameter uncertainties (iii) reflect theuncertainty intrinsically associated with life cycle inventory data. Uncertaintyanalysis was performed by using MonteCarlo analysis (Tonini et al., (II )). Thisshould be done by comparing two selected LCA scenarios in each run of theMonteCarlo analysis so to take into account the ‘correlated’ uncertainties (i.e. parameters uncertainties which are present in both LCA scenarios). Sensitivityanalysis was instead performed by changing individual assumptions or parameters (as assumed in the baseline) and then comparing the ‘new’ LCAresults obtained with the baselines. This analysis was applied to all the LCAstudies performed within this thesis.



2 1

T a b

l e 2

. O v e r v i e w o f

t h e r e l e v a n t m e t h o d o l o g i c a l a s s u m p t i o n s

i n t h e

L C A p a p e r s

f o r m i n g t h e b a s i s

f o r t h i s t h e s i s . S A : s e n s i t i v i t y a n a l y s i s ;

U A :

u n c e r t a i n t y a n a l y s i s ;

E : e l e c t r i c i t y ; H :

h e a t ; C H P ( D H ) : d i s t r i c t h e a t i n g p r o v i d e d

b y a

l o c a l C H P p l a n t . [ 1 ] :

S t - C l a i r e t a l .

( 2 0 0 8 ) ; [ 2 ] : H

a m e l i n e t a l .

( 2 0 1 2 ) ; [ 3 ] : S c h m i d t ( 2 0 0 8 ) ; [ 4 ] : S c h m

i d t ( 2 0 0 7 ) .

P a p e r

F u n c t i o n a l u n i t

T e m p o r a l s c o p e

T e c h n o l o g i c a l

s c o p e

G e o g r a p h i c a l

s c o p e

M a r g i n a l s

d L U C

i L U C

S A

U A

I

P r o v i s i o n o f p r i m a r y e n e r g y t o

s a t i s f y s o c i e t y n e e d s

2 0 3 0 a n d

2 0 5 0

F u t u r e

t e c h n o l o g i e s

D e n m a r k

C r o p : s p r i n g b a r l e y

[ 1 ] , [ 2 ] , [ 4 ]

[ 3 ] , [ 4 ]

X

I I

B i o e n e r g y p r o d u c t i o n

f r o m 1 h a

o f D a n i s h a r a b l e

l a n d

2 0 1 2 - 2 0 3 2

S t a t e o f

t h e a r t

( e x i s t i n g )


D e n m a r k

E : C O ( S A : N G

)

H : N

G ( S A : C O

)

C r o p : s p r i n g b a r l e y

[ 2 ]

S e c t i o n 3 . 3

X

X

I I I

T r e a t m e n t o f o n e

t o f r M S W

2 0 1 2 ( p r e s e n t )

S t a t e o f

t h e a r t

( e x i s t i n g )


D e n m a r k

E : C O

, N G

H : C

O , N

G

N o t r e l e v a n t


X

V


t o f r M S W

2 0 1 2 ( p r e s e n t )

S t a t e o f

t h e a r t

( e x i s t i n g )


S p a i n

E : C O

, N G

H : n o t r e l e v a n t



X

V I


t o f M S W

2 0 1 5 - 2 0 3 0

F u t u r e


E u r o p e

E : C O

, N G

H : C

O , N

G , C H

P ( D H )



X

X



22



23

4. Processes in bioenergy and WtE system- Inventory data

This section describes the relevant processes involved in bioenergy and WtEsystems.

4.1 Agricultural processesWhen assessing bioenergy systems the choice of the agricultural inventory iscrucial as it determines dLUC impacts and later energy production that dependsupon the crop yield. Critical emissions occurring during agricultural processesare CH4, N2O and NO3

- emissions. These are directly related to the amount of

mineral and organic N-fertilizers utilized. From this perspective perennial energycrops have significantly lower requirements than annuals (Hamelin et al., 2012).Further, they also have higher yields along with other correlated benefits such asless soil disturbance and increase in soil organic carbon (SOC). Overall, There iswide consensus on that perennial energy crops are currently the most efficientand sustainable feedstock for the purpose of bioenergy in temperate climates(Bessou et al., 2011, Dauber et al., 2010, Valentine et al., 2012).

The crop yield plays a significant role as it determines the energy production at alater stage (as earlier mentioned). Crops yields vary depending upon type ofcrops and geographical conditions (climate). For example ryegrass yield inDenmark is typically between 9 and 18 t DM ha-1 (Moeller et al., 2000). Theyield for willow is estimated to ca. 9-17 t DM ha-1 after Hamelin et al. (2012).The average yield for Miscanthus in DK is ca. 15 (autumn harvest) and 10(spring harvest) t DM ha-1 (Hamelin et al., 2012). Corresponding Miscanthus yields in Central and Southern Europe may be significantly higher; for instanceLewandowski et al. (2000) reported autumn yields of 25 t DM ha-1 from trials inGermany. It should be noted how some crops may achieve different yieldsdepending on the harvesting period (e.g. Miscanthus ). This also induces adifferent carbon sequestration in the soil because of the variation in the amountof above- and below-ground residues (Hamelin et al., 2012).

4.2 Storage processes Storage is needed within the bioenergy chain as biomasses accumulate seasonallyand the energy plants have, instead, to be fed and run continuously. In addition, biomass prices will be market-driven and the producers will sell the crops



24

whenever the prices will be convenient, therefore storage will be likely tohappen. Mass losses during this stage might decrease the final energy recovery aswell as induce emissions of CH4 and N2O thus reducing the overall GHG

savings. For example when Danish ryegrass is harvested (beginning of summer)the water content is about 80% (Hamelin et al., 2012). This means that wetstorage might be possible in the form of ensiling, whereas dry storage wouldrequire prior drying. The drying process for ryegrass is typically operated onfield and related mass losses may be in the range 10-30% of the initial DMcontent (caused by microbial respiration, precipitations as well as by the differentoperations such as turning, mowing and baling) (Mcgechan, 1989, Prochnow etal., 2009). Subsequent indoor dry storage typically leads to losses between 1.1%and 11% similarly to dry lignocellulosic (woody-like) biomasses (Emery andMosier, 2012). According to the same authors, if wet storage (i.e. ensiling) is thechoice, the mass losses may be as high as 20-25% for ryegrass with water contentabove 80%.

For the case of willow (about 50% water content at harvest) and woody biomassdifferent techniques for natural drying exist. These are typically performedduring the storage period leading to mass losses estimated between 3.5% and6.1% for rods (Gigler et al., 2004, Kofman and Spinelli, 1997, Jirjis, 1995).Thermal drying, although possible, is associated with significant economical andenergy costs which make it less attractive (Lewandowski and Heinz, 2003). No-drying also represents an alternative; however, wet willow (in form of chips) was proven to determine high dry matter losses due to increased microbial activityand degradation (Kofman and Spinelli, 1997, Jirjis, 1995, Wihersaari, 2005).Further, it should be noted that when co-digesting energy crops with manure, theenergy production per unit-input increases with the dry matter content of the co-substrate (Tonini et al. (II )). Therefore, dry co-substrates are favourable over wetones to maximize the energy production. In the case of MSW, the storage

processes were not considered as relevant within this thesis. The storage processes were assumed equal for all the management scenarios assessed as thesewere not subject to the same assumptions as for biomass systems (i.e. pricesdependency).



25

4.3 Pre-treatment processes 4.3.1 Pre-treatments for biomass

When anaerobic digestion of lignocellulosic biomasses is considered (e.g. willowand Miscanthus ) a pre-treatment is needed to enhance the susceptibility to biodegradation of lignocellulose. This includes size comminution (shredding)and further enhancement treatment. These processes need to be accounted for asincreased energy consumptions of the system under assessment, generally in theform of heat and electricity (additional fuels and chemicals might also berequired). Mass losses (about 10%) can also occur if strong oxidizing agents areused to enhance lignin biodegradation (Bruni et al., 2010). Different pre-treatments exist; these can be classified into the following: i) biological

(enzymatic), ii) chemical, iii) mechanical and iv) hydrothermal (Bruni et al.,2010). Combinations of these are also possible.

In the case of thermal energy conversion (gasification, direct combustion and co-firing), different pre-treatments may be required depending on the type oftechnology; for instance, many of the Danish small-scale biomass combustionCHP plants have been adapted to minimize pre-treatments and energyconsumptions. In these facilities, pelletization, size comminution (shredding) and pulverization are generally not a requirement. When direct co-firing is applied, prior pelletization and milling are instead performed (generally the pellets aremilled along with coal and combusted together). This is not the case for parallelco-firing (i.e. independent biomass boiler; the steam from coal and biomass arethen mixed and sent to the turbines). In the latter case, in fact, size-comminutionis performed but pelletization and milling is generally not a requirement. Finally,in the case of gasification, the pre-treatment varies upon the technology(fluidized bed, fixed bed, etc.). Gasification in fluidized bed typically requires biomass comminution (10-50 mm) and drying (water content is recommended

below 20%) (Hughes and Larson, 1998). The main environmental emissionsassociated with pre-treatments are connected to fuel and energy provision for theoperations.

4.3.2 Pre-treatments for MSWWhen MSW (or rMSW) is considered, different pre-treatments prior tosubsequent energy recovery (or eventually disposal) are possible. These include:i) source-segregation, ii) mechanical pre-treatments prior to anaerobic digestion,iii) material recovery in dedicated selection plants (i.e. MRF) and innovative



26

technologies such as iv) waste refining and v) autoclaving. Notice that if MSW issent directly to incineration (without prior segregation or treatment) a pre-treatment consisting on simple removal and/or shredding of bulky elements

might be needed (Hulgaard and Vehlow, 2010).

Source-segregation (i) is possible and different techniques exist for separation ofselected waste material fractions (Christensen and Matsufuji, 2010). Accordingto the same authors the efficiency varies dramatically depending on the collectionscheme: from 10-20% for collection centres to 60-90% for full-service collection(door to door). Source-segregation is particularly relevant for organic (vegetablewaste, animal waste and tissues) in the perspective of reducing the amountdisposed of in landfill as enforced by the EU directive (CEC, 1999). To thisrespect however, the results are often far from the expectations as segregation atthe household and further mechanical pre-treatments prior to biologicalconversion may lead to significant mass losses. Although source-segregationmay achieve efficiencies between 60% and 80% for full-service collection(Christensen and Matsufuji, 2010), recent studies found efficiencies as low as ca.22-26% (Bernstad, 2012).

When organic waste is source-segregated further mechanical pre-treatment (ii) prior to digestion is required (Jansen, 2010). The function is to remove unwanteditems and achieve size reduction of the substrate. Typically the pre-treatmentconsists of: shredding (with bags openers), metals removal (with magnets),sieving for plastic removal (with disc sieves or trommels; gravity separation in pulpers is also an option for wet processes). Additionally, hygienizationtreatment might also be required. The residues consist of dirty plastic bags(typically incinerated), metals (highly contaminated with organic) as well asother heavy materials such as stones and glass (generally landfilled). Such a pre-treatment induces additional mass losses of organic waste: Bernstad et al.

(2012), for example, found an average mass loss of about 20% (range 2-45%) ofthe incoming organic waste (corresponding to 13-39% on a DM basis). Based onthis information and efficiencies, Tonini et al. (VI ) investigated a number ofscenarios involving source-segregation of materials and organic waste. Thesewere compared with other scenarios not involving segregation in order toevaluate the potential benefits and drawbacks of each strategy.

MRF (iii) generally indicates any mechanical-treatment facility aiming atrecovering selected materials from the waste. MRFs can be classified as follows(Christensen and Bilitewski, 2010): i) single MRF, upgrading a single segregated



27

material fraction; ii) commingled MRF, sorting a commingled collected fractionconsisting of more than one waste material fraction; iii) mixed MRF, sortingrMSW; iv) MBT, sorting a range of materials from the mixed waste (generally

rMSW) and using biological treatment to stabilize the organic fraction of MSW(OFMSW). The type of MRFs used in a waste management system is related tothe collection scheme. The efficiency of materials recovery ranges from 60% to98% for manual sorting of source-segregated waste (typically 90%) and from50% to 98% (typically 75-85%) for mechanical separation of commingled wastematerial fractions (Tchobanoglous and Kreith, 2002).

Mechanical-biological treatment is a type of MRF aiming at recovering materialsand energy from the mixed waste by using a combination of mechanical and biological operations. Two types exist: A) mechanical-biological stabilization(MBS) or biodrying, which first composts the waste for drying prior to extractionof a larger RDF fraction and B) mechanical-biological pretreatment (MBP),where the organic fraction is separated and biologically stabilized prior tolandfilling and recyclables as well as RDF are recovered from the residual coarsefraction. MBP aims at stabilizing the organic to minimize gas as well as leachateemissions in landfill while MBS maximizes RDF recovery. Within this generalclassification, multiple variations can be found and it can be stated that probablythere are no two identical plants (Bilitewski et al., 2010). In the type A theorganic fraction is dried and sent to combustion along with plastic, paper andhigh-calorific value materials. In the type B the organic fraction is separated andsent to anaerobic digestion (and/or composting) for energy recovery andstabilization. The stabilized organic material (namely compost) is generallylandfilled or used as landfill daily cover, as its poor quality (mainly related tohigh metals content) does not allow for use on land (Montejo et al., 2010).Montejo et al. (V) performed an assessment of eight management strategies based on MBT in Castilla y Leon (Spain). This also included waste

characterization analyses specific for each individual plant.

The waste refinery process (iv) aims at generating two products from theincoming mixed MSW (Figure 3): i) a bioliquid (i.e. slurry composed ofenzymatically liquefied organic, paper and cardboard) and a solid fraction (i.e.non-degradable waste materials). The refinery process consists of two main sub- processes, i.e. heating and enzymatic treatment. A detailed description of theenzymatic processing can be found in Jensen et al. (2010). The bioliquid can beexploited for biogas production, co-combusted in coal-fired power plant orutilized for producing ethanol. This, compared with direct incineration, provides



28

additional flexibility to the energy system as the energy production could beregulated and storage possible in form of bioliquid/biogas. This may be relevantin the perspective of energy systems having high penetration of windenergy and

other fluctuating renewables as illustrated in previous studies (Lund, 2010, Lundand Mathiesen, 2009, Mathiesen et al., 2011, Mathiesen et al., 2011). The solidfraction can be further treated to separate and recover valuable materials such asmetals and plastic. The remaining residual solid (mainly non-recyclable plastic,textiles, yard waste, undegraded organics and glass pieces) can be combusted forenergy recovery. A pilot-plant waste refinery established in Copenhagen (DK)has been investigated within this thesis (Tonini and Astrup, III , Tonini et al., IV ,Tonini et al., VI ). This also included a waste characterization study performedwithin Tonini et al. ( IV ).

Autoclaving (v) is a hydrothermal process occurring in wet environmentalconditions with high temperature and pressure provided by saturated steam(Stentiford et al., 2010). The result of autoclaving is a reduction of the initialvolume of the input waste (corresponding to ca. 80%), sterilization of pathogens,removal of liquids, compaction of the plastics and removal of labels on glass and

plastic containers. In addition, all the biodegradable waste material fractions(mainly paper, cardboard and organic matter) are combined into a single productnamely organic fiber. This can be further treated with anaerobic or aerobicdigestion process to recover energy and stabilize the materials. Recyclables suchas metals, plastic and glass may be sorted from the remaining solid output.

Figure 3. Illustration of the waste refinery process (after Tonini et al., III ): bioliquid andresidual solid are sent to energy conversion. Metals are sent to recycling. Selected plasticfractions can be sorted out from the solid fraction within the refinery process and sent torecycling (not visualized).



29

4.4 Energy conversion technologiesThe energy conversion technologies for biomass and waste can be generallyclassified into biological (e.g. anaerobic digestion) and thermal (pyrolisis,gasification and combustion). Landfilling may also recover energy in the form of biogas; in this perspective the landfill body acts as an anaerobic reactor wherewaste and precipitation are the inputs and leachate and gas the outputs(Christensen et al., 2010, Willumsen and Barlaz, 2010). An overview of relevantair emissions and energy efficiencies for a number of conversion technologies is presented in Table 3-4.

4.4.1 Anaerobic digestion

Anaerobic digestion technologies (also called biogasification) can be classified aswet/dry, mesophilic/termophilic, one-stage/two-stage, one-phase/two-phase(Jansen, 2010). Anaerobic digestion produces two products from the input: biogas and digestate (plus liquid from dewatering if implemented). For organicwaste both dry mono-digestion and wet co-digestion (e.g. with municipalwastewater or animal manure) are applied. Dry digestion has the advantage ofrequiring less digestion volumes; this maximizes the specific energy production(per unit of reactor, or unit of input wet basis). Co-digestion is used to balancenutrients content and/or to boost the energy production of selected substrates.

For agricultural biomasses with high lignocellulose content (e.g. willow and Miscanthus ) mono-digestion may encounter problems and finally failures due tosub-optimal macro- (e.g. unfavorable C/N ratio) and micro-nutrients content. Inthe light of this, co-digestion with animal manure (or OFMSW) may be asolution (Nges et al., 2012, Alvarez et al., 2010, Mshandete et al., 2004). Manurerepresents an important energy resource (see section 1), that is, in the case ofDenmark largely unexploited. The reason for this is the scarce economical and

technical attractiveness of manure mono-digestion due to its low DM content (2-10%) inducing low specific energy production (per unit of reactor); therefore, thecurrent management in Denmark is represented by storage and further use onland. This is responsible for significant environmental impacts due to emissionsof CH4, N2O and NO3

- (Hamelin et al., 2011). A detailed mass-balance approachto model co-digestion of animal manure and energy crops is presented in Toniniet al. (II ). The main environmental emissions associated with the digestion process are connected to fuel and energy provision for the operations and CH4 leakages from the reactor. These may vary from 0% to 10% of the CH4 produced

(Eggleston et al., 2006). However, recent LCA studies tend to use 1% for



30

assessing state-of-the-art or future technologies as the insulation of the reactorshas significantly been improved (Hamelin et al., 2011, Boerjesson and Berglund,2006, Jungbluth et al., 2007a).

The produced biogas can be used in gas engines, gas turbines, boilers, co-fired in power plants or upgraded to transport fuel (95% methane content, v/v) or tonatural gas quality. In most of the cases combustion in gas engines is performed.The net electricity efficiency varies between 34% and 42% relative to the LHVof the input-gas (Fichtner, 2004). The overall net energy recovery can reach 95%(without flue-gas condensation) and 103% (with flue-gas condensation)according to Energistyrelsen (2012). For the specific case of gas engines, therelevant environmental emissions (Table 3) are NOx, SO2 (especially for biogasfrom MSW), CO and uncombusted CH4 (Nielsen et al., 2010).

Table 3 . Selected air emissions from biomass and bio/syngas combustion (Nielsen et al., 2010).Values are expressed per GJ of primary energy (LHVwb, i.e. LHV wet basis) of the fuelcombusted. PCDD/F-: dioxins and furans (as Polychlorinated Dibenzo-p-dioxins, i.e. PCDDs);TSP: total suspended particulate; UHC: unburned hydrocarbons.

Air emission Unit Biogasengines

Syngasengines

Strawcombustion

Woodcombustion

CO g GJ-1 310 586 67 90CH4 g GJ-1 434 13 <0.47 <3.1

N2O g GJ-1 1.6 2.7 1.1 0.83 NOx g GJ-1 202α 173 125 81

PCDD/F- ng GJ-1 <0.96 <1.7 <19 <14

HCl g GJ-1 - - 56 -

Naphthalene μg GJ-1 4577 8492 12088 2314 NMVOC g GJ-1 10 2.3 <0.78 <5.1

∑PAH μg GJ-1 <606 <181 <5946 <664

SO2 g GJ-1 - - 49 <1.9

TSP g GJ-1 - - <2.3 10

UHC g GJ-1 333 12 <0.94 <6.1α NOx can be reduced to 60 g GJ-1 provided installation of SCR (Energistyrelsen, 2012).

4.4.2 Pyrolisis and gasificationPyrolysis and gasi cation thermally convert carbon-containing substrates into arange of products (syngas, char, coke, ash, and tar) in a not-completely oxidizingenvironment. Overall, pyrolysis generates products such as syngas, tar, and char,while gasi cation maximizes the conversion of the carbon-containing substrates

into syngas for further use. The output composition and relative amounts of the



31

products are largely dependent upon the inputs and the process con guration. Forgasification, important parameters describing the efficiency of the process areCGE (cold gas efficiency) and CCE (carbon conversion efficiency). The CGE

defines the fraction of the feedstock chemical energy (as LHV, dry basis)remaining in the syngas (and not lost as, e.g. heat or in the residue). It isexpressed as the ratio between the amount of energy in the syngas (after gascleaning) and the amount of energy in the biomass (as LHV, dry basis). The CCEdefines the proportion of the feedstock C that is transferred to the syngas. InTonini and Astrup (I) and Tonini et al. (II ) a review of literature values for CGEand CCE for a number of different biomasses is presented. Generally, CGEvaries between 55% and 85% for fluidized bed reactors, whereas CCE between91% and 99%. However, progresses in the technological development may leadto improved CGE: for example, Ahrenfeldt et al. (2006) reported a CGE of 93%for woodchips conversion in a two-stage fixed bed gasifier; Arena et al. 2010reported a CGE up to 94% for selected plastic materials in a pilot-scale fluidized bed reactor. The main environmental emissions are associated to energy and fuel provision for the operations. The produced residues may be very stable and solid(similar to vitrified or melted residues from incineration) especially in the case ofhigh-temperature processes (Astrup and Bilitewski, 2010).

As for the biogas, the syngas produced can be combusted in gas engines, gasturbines, boilers, co-fired in power plants or upgraded to transport fuel (95%methane, v/v) or to natural gas quality. The combustion efficiency is similar to biogas, although additional purification from tars and impurities might be neededfor some applications, e.g. turbines (thus potentially reducing the overallefficiency) (Ahrenfeldt et al., 2006, Astrup and Bilitewski, 2010, Arena et al.,2010, Arena et al., 2011). The overall net biomass- or waste-to-electricityefficiency is estimated to ca. 10-20% for steam turbines (Astrup and Bilitewski,2010, Arena et al., 2010), 13-28% for gas engines (Ahrenfeldt et al., 2006, Arena

et al., 2010) and 15-25% for gas turbines (Arena et al., 2010). The relevantenvironmental emissions (Table 3) are NOx, SO2 (especially for syngas fromMSW), N2O and CO (Nielsen et al., 2010). Emissions of CH4 from syngascombustion in gas engines are significantly decreased compared with biogas.

4.4.3 Direct combustion and co-firing of biomassCombustion of biomasses and waste may be performed in a variety of plants andconfigurations. Direct combustion is typically performed in small and mediumscale combustion and incineration plants (see Tonini et al.,II ). Co-firing is,



32

instead, generally performed in large coal-fired power plants. Different types ofco-firing configurations exist: direct (in the coal boiler), parallel (independent biomass and waste boiler) and indirect (biomass or waste are gasified to syngas

that is later combusted in the coal boiler or in an independent one). For biomass,the combustion efficiency is often determined by the scale of the plant. However,the fuel property also affects the overall efficiency. In Tonini et al. (II ) a reviewof existing biomass combustion CHP plants is reported. The net electricityefficiency for recently commissioned small-to-medium scale biomass CHP plantsvaries between 25% and 29%, relative to the LHV of the input-fuel. Co-firing inlarge scale coal-fired CHP plants achieves higher efficiency (35-41%). Forexample, the power plant Avedøreværket in Copenhagen may achieve (full-load)a net electricity efficiency of 41% in CHP mode and up to 49% in condensingmode (DONG, 2009). Relevant environmental emissions of biomass combustion plants are NOx and SO2 (Table 3). Ash residues are generally applied on land orlandfilled; re-use for road construction is also an option.

4.4.4 Incineration and co-firing of MSWMSW incineration is largely practiced in Denmark and Europe. In 2009 about20% of the MSW produced in EU was incinerated (Eurostat, 2011). Three mainconfigurations exist: moving grate, rotary kiln and fluidized bed (Hulgaard and

Vehlow, 2010). Moving grate is the most common technology. The total grossenergy recovery for state of the art plants can be as high as 103% relative to theLHV (wet basis) provided flue-gas condensation. The gross electricity efficiencymay range between 25% (CHP mode) and 30% (condensing plants such as AfvalEnergie Bedrijf in Amsterdam) relative to the LHV (wet basis) of the waste-input. This corresponds to net electricity efficiency of ca. 22-26% and net total ofca. 97-99% (relative to LHV) (Energistyrelsen, 2012). Older plants typicallyhave lower performances: a survey of 231 EU waste incinerators found anaverage electricity recovery of 20.7% in condensing mode and 14% in CHPmode (Reimann, 2009).

The main environmental impacts of incineration are related to: i) fuel and energy provision for the operations, ii) air emissions and iii) residues. Electricity provision typically corresponds to ca. 10% of the total energy produced(Reimann, 2006). Air emissions for SOx, NOx, HCl, dioxins and, to some extent,Hg are directly related to the technological level of the flue-gas cleaning(Damgaard et al., 2010). The emissions of heavy metals are, instead, primarily

determined by the waste composition (Astrup et al., 2011). The residues (iii) are



33

represented by bottom and APC residues (mainly fly ash). Bottom ash can belandfilled in inert landfills or re-used for road construction after metals extraction(Birgisdottir et al., 2007). Fly ash can be used as back-filling material for old salt

mines, landfilled in appropriate hazardous waste landfills or used as neutralizingagent in cement production (Fruergaard et al., 2010b).

Co-firing of selected MSW fractions (e.g. RDF and solid recovered fuel, i.e.SRF) in a variety of combustion plants (e.g. coal fired power plants, biomass boilers, cement kilns, etc.) is also an option. If co-firing of RDF or SRF in coalfired power plants is applied, the net electricity efficiency may raise to 35-41%relative to the LHV (wet basis) thanks to the higher steam and pressure parameters achieved in the boiler. The main environmental concern of co-firingis related to the increased air emissions (from the waste input) because of limitedair pollution control (Fruergaard and Astrup, 2011, Rechberger, 2010).

4.4.5 Liquid biofuelsBiodiesel can be produced via two main pathways: 1) transesterification of oil plants (rapeseed or palm) producing RME-biodiesel and 2) gasification oflignocellulosic biomass (e.g. wood) followed by Fischer and Tropsch process producing FT-biodiesel. The transesterification from 1 t DM rapeseed produces:

RME (0.35 t), rape meal (0.6 t) and glycerine (0.038 t). The gross energyefficiency of biomass to biofuel is about 63% (this value does not account forenergy consumption of the process and for byproducts).

The thermochemical conversion uses a first gasification step to generate syngaswhich is then upgraded to biodiesel through Fischer and Tropsch process. Thegross energy efficiency of biomass to biofuel is about 45% (Jungbluth et al.,2007b). Additional energy consumptions for the Fischer and Tropsch processdecrease the energy efficiency to ca. 40% (net) (Jungbluth et al., 2007b).Research and inventory data on the process are still at an early stage.

Bioethanol can be produced from a variety of carbohydrates-rich substrates (forexample, grass, wood, molasses, wheat, maize, straw, etc.). The current focus ofthe research is on conversion of residual agricultural products such as straw, asthese do not involve iLUC. The conversion of straw into ethanol (Larsen et al.,2008, Bentsen et al., 2009) produces the following outputs per 1 t DM straw:ethanol (0.21 t), C5 molasses (0.25 t, 30% water content) and solid biofuel (0.35

t, 10% water content).



3 4

T a b

l e 4

. O v e r v i e w o f s e l e c t e d

b i o m a s s a n d

W t E c o n v e r s i o n p r o c e s s e s .

R e l e v a n t e n v i r o n m e n t a l e m i s s

i o n s a r e a l s o r e p o r t e d .

C D : c o n d e n s i n g m o d e ;

C H P : c o m

b i n e d h e a t a n d p o w e r ;

F G C :

f l u e - g a s c o n d e n s a t i o n ; η e l : n e t e l e c t r i c i t y e f f i c i e n c y ; η t o t :

n e t t o t a l e f f i c i e n c y ; n . r . : n o t r e p o r t e d .

[ 1 ] :

E n e r g i s t y r e l s e n

( 2 0 1 2 ) ; [ 2 ] : T o n i n i e

t a l . ( I I ) ; [ 3 ] : D O N G ( 2 0 0 9 ) ; [ 4 ] : F i c h t n e r ( 2 0 0 4 ) ; [ 5 ] : A r e n a e t a l .

( 2 0 1 0 ) . B i o m a s s c o - f i r i n g

i s b a s e d o n

t h e d a t a

f r o m

t w o

D a n i s h p l a n t s

( A v e d ø r e v æ r k e t a n d O e s t k r a f t ) . W a s t e c o - f i r i n g

i s a s s u m e d

t o h a v e s i m

i l a r e f f

i c i e n c i e s a s

f o r b i o m a s s c o - f i r i n g

( R e c h b e r g e r ,

2 0 1 0 ) .

E n e r g y c o n v e r s i o n p r o c e s s

η e l

η t o t

E n v i r o n m e n t a l e m

i s s i o n s

S o u r c e

N o t e

B i o m a s s c o m

b u s t i o n ( s m a l l - t o - m e d i u m

s c a l e

C H P p l a n t s )

2 5 - 2 9 %

( 4 4 - 4 6 % C D )

9 3 - 9 5 %

( 1 0 3 - 1 0 6 % F G C )

N O

x , S O 2 , C

O 2 , N 2 O , d u

s t

[ 1 ] , [ 2 ]

S m a l l :

1 - 1 0

M W

M e d i u m : 1

0 - 1 0 0 M W

B i o m a s s c o - f i r i n g

( l a r g e s c a l e

C H P

p l a n t s )

3 5 - 4 1 %

( 4 4 - 4 9 % C D )

9 3 - 9 5 %

( 1 0 3 - 1 0 6 % F G C )

N O

x , S O 2 , C

O 2 , N 2 O , d u

s t

[ 1 ] , [ 2 ] , [ 3 ] , [ 4 ]

η t o t

i s a s s u m e d s i m

i l a r t o

s m a l l a n d m e d i u m

C H P

B i o g a s / s y n g a s c o m

b u s t i o n

i n

G E

( C H P )

4 0 - 4 8 %

( 3 4 - 4 2 % ) *

8 8 - 9 6 %

( 1 0 3 % F G C )

N O

x , S O 2 , C

O 2 , N 2 O , C H 4

[ 1 ] , [ 4 ]

* [ 4 ] s u g g e s t s a

l o w e r

r a n g e f o r

t h e

η e l

e f f i c i e n c y

B i o m a s s

t o

s y n g a s

a n d

s y n g a s

c o m b u s t i o n i n s t e a m c y c l e s

1 0 - 2 0 %

n . r .

N O

x , C O

2

[ 5 ]

o v e r a l l n e t η

e l b i o m a s s - t o -

e l e c t r i c i t y

B i o m a s s

t o

s y n g a s

a n d

s y n g a s

c o m b u s t i o n i n g a s

t u r b i n e s

1 5 - 2 5 %

n . r .

N O

x , C O

2

[ 5 ]

o v e r a l l n e t η

e l b i o m a s s - t o -

e l e c t r i c i t y

W a s t e i n c i n e r a t i o n ( C H P )

2 5 - 3 0 % *

1 0 0 - 1 0 3 % F G C α

N O

x , S O 2 , C

O 2 , H

C l ,

d i o x i n s , m e t a l s ,

d u s t

[ 1 ]

* g r o s s e f f i c i e n c y

f o r s t a t e

f o r t h e a r t

W t E p l a n t s

W a s t e c o - f i r i n g

( C H P )

3 5 - 4 1 %

( 4 4 - 4 9 % C D )

9 3 - 9 5 %

( 1 0 3 - 1 0 6 % F G C ) β

N O

x , S O 2 , C

O 2 , H

C l ,

d i o x i n s , m e t a l s ,

d u s t

[ 2 ] , [ 4 ]

E f f i c i e n c y i s a s s u m e d a s

f o r b i o m a s s

c o - f i r i n g

α T y

p i c a l l y t o t a l e f f i c i e n c i e s w

i t h F G C m a y

b e a s

h i g h a s

1 0 3 % p r o v i d e d

t h a t N O x r e m o v a l

i s o p e r a t e d w

i t h S N C R

. I f S C R

i s i n s t e a d o p e r a t e d ,

t h e e l e c t r i c i t y e f f i c i e n c y

i s s l i g h t l y

d e c r e a s e d a n d

t h e t o t a l e f f i c i e n c y i s r e d u c e d b y c a .

2 . 5 % c o m p a r e d w

i t h t h e m a x i m u m p o t e n t i a l ( t h a t i s c a .

1 0 3 % ) . β S a m e a s f o r w a s t e

i n c i n e r a t i o n .



35

5. Environmental performance of bioenergyand WtE systemsThis section highlights the findings of a number of case studies assessing theenvironmental consequences associated with energy production from biomassand waste. The section elaborates on the results of Tonini and Astrup (I), Toniniet al. (II ), Tonini and Astrup (III), Tonini et al. (IV ), Tonini et al. (VI ) andMontejo et al. (V).

5.1 Case study: future energy scenarios for DK5.1.1 Modeling aspectsIn Tonini and Astrup (I) a number of future energy scenarios for Denmark (onefor 2030 and three for 2050) were compared with the 2008 Danish energy systemused as reference. The ambition of Denmark is to achieve a 100% renewableenergy system by 2050 (Lund and Mathiesen, 2009, Mathiesen et al., 2011,Danish Ministry of Climate, Energy and Buildings, 2011). In order to do so, anumber of technical measures should be implemented. This primarily translatesinto improving the efficiency of power plants, electricity transmission,transportation, district heating networks, building insulation, etc., in order to

reduce the energy demand and consequent primary supply. Further, wind andhydro power, photovoltaic and geothermal should be significantly increased(among the others: Mathiesen et al., 2011, Mathiesen et al., 2009).

Energy system analysis (ESA) was used to design a number of future scenariosunder technical constraints represented by technologies and capacity ofinterconnectors. For Denmark, a significant increase in wind energy penetrationwas recommended (Mathiesen et al., 2011). This, however, needs solid fuelsand/or storable energy carriers (e.g. biomass, fossil fuels, biogas, syngas, etc.) to

balance the associated intermittent production. This leads to that ca. 35-50% ofthe primary energy supply should be covered by biomass and storable carriers(Lund, 2010, Lund and Mathiesen, 2009, Lund, 2007, Mathiesen et al., 2011). Inaddition, based on the same energy analyses, storable fuels are still required for part of the transport sector (e.g. heavy vehicles, ships, defence, etc.) even thoughelectrification of passenger vehicles is envisioned. The environmental profile ofthese energy scenarios was evaluated by means of LCA.



36

Figure 4 . Illustration of the energy flows (PJ) for the three assessed scenarios ‘2050’ (fromTonini and Astrup,I). The scenarios differ for how the energy needed in the transport sector is provided. GH: geothermal heating, ST: solar heating, PV: photovoltaic, Hydro: hydropower, NG: natural gas, Syn: syngas, BG: biogas, BP: byproducts, E: electricity, H: heat, H*: processheat (industry), L: losses, DS: diesel, BD: biodiesel, AF: aviation fuel, SF: synthetic fuel.



37

Figure 4 shows possible energy systems for Denmark in 2050: three 2050scenarios were evaluated; these differ for how the fuel needed for the transportsector was provided. Three alternatives were considered: 1) fossil diesel

(scenario ‘2050 CSV’), 2) RME- (rape methyl ester) biodiesel via rapeseedtransesterification (scenario ‘2050 RME’) and 3) FT- (Fischer and Tropsch) biodiesel via willow gasification and further FT processing (scenario ‘2050BtL’). Note that oil is still present in all the addressed scenarios as domestic andinternational aviation was assumed to rely on fossil fuels (primarily jet fuel andkerosene). To date, the research on aviation biofuels is still at an early stage andvery little information is available on processes and efficiencies.

5.1.2 Key resultsThe analysis revealed that, even assuming future optimistic decreases in theenergy demand, the domestic available biomass resource potential (ca. 182 PJ, asestimated in this study) was not sufficient to cover the projected primary energysupply (between 559 and 588 PJ). However, the results also indicated thatsignificant GHG emission reductions may be achieved by a combination of i)reduced energy demand (hence supply), ii) increased share of windenergy and iii)replacement of fossil fuels with domestically available biomass and, to the extentneeded, energy crops.

While most of the required electricity and heat could be provided by integratingwindenergy and highly efficient use of bio/syngas (e.g. in fuel cells) fromdomestically available biomass, the bottleneck lied in the provision of diesel-like biofuels for heavy terrestrial transport, ships and defence. With respect to this,extraction and production of oil or, alternatively, cultivation of specific energycrops was necessary in order to supply diesel-like fuel. RME-biodiesel was theworst option for all the environmental categories. The overall GHG emissionequalled ca. 287 Gg CO2-eq. PJ-1 fuel (Gg CO2-eq. PJ-1). This was far higher thanfor fossil diesel (89 Gg CO2-eq. PJ-1 fuel). FT-biodiesel showed practically noGHG savings compared with fossil diesel when possible benefits (highlyuncertain) associated with residual biochar from gasification and FT processingwere not considered. An overview of the GHG impacts associated with the biofuels is reported in Table 7 along with the findings from other studies.

Significant aquatic eutrophication effects were induced by rapeseed cultivationAdditionally, NOx tailpipe emissions from biodiesel combustion in car engines

(both RME- and FT-biodiesel) are significantly higher than the corresponding



38

emissions from fossil diesel combustion, as highlighted by a number of studies(among the others: Zhu et al., 2010, Sun et al., 2010, Mazzoleni et al., 2007,Wang et al., 2000). However, it may be envisioned that future development in the

exhaust gas cleaning technology may reduce this impact. Land occupation wasalso subject to significant increases in all the future scenarios compared with thereference (2008) due to increased crops cultivation.

The main recommendation that can be drawn from this case study is that in orderto minimize environmental impacts it is favourable to focus on the optimizationof bioenergy production from residual domestically available biomass resources(such as MSW, animal manure, straw, grasses, forest residues, wood residues,etc.) rather than on energy crops. Additionally, production of RME-biodieselshould be avoided and alternative diesel-like biofuels should be encouraged.

5.2 Case study: bioenergy from perennial crops 5.2.1 Modeling aspectsAs aforementioned (section 5.1) for the case of Denmark the domesticallyavailable biomass resources (along with windenergy and other renewables) arenot sufficient to satisfy the energy demand even when this is assumedsignificantly reduced compared with today’s. Therefore, if the ambition is toreach a 100% renewable energy system, energy crops cultivation may be needed.

In Tonini et al. (II ) a number of bioenergy scenarios were assessed by means ofLCA. The scenarios included three perennial crops namely i) ryegrass ( Lolium

perenne ), ii) willow (Salix viminalis ) and iii) Miscanthus giganteus. Four biomass-to-energy conversion technologies were modeled: I) anaerobic co-digestion of biomass with raw pig manure, II) thermal gasification, III) direct biomass combustion in small-to-medium scale biomass CHP plants and IV) co-

firing in large scale coal-fired CHP plants. The total added up to 3 x 4 = 12 bioenergy scenarios. The assessment included quantification of LUC, C, N andenergy flow analysis as well as sensitivity and uncertainty analysis on therelevant parameters and scenario uncertainties (see also section 3.4). An exampleof carbon flow chart is presented in Figure 5 for the case of anaerobic co-digestion of willow with raw pig manure.



39

Figure 5 . Illustration of the (biogenic) carbon flows for the case of co-digestion of willow withraw pig manure (from Tonini et al.,II ). Fossil C flows associated with energy and fuelconsumptions for machinery and operations throughout the whole bioenergy chain are notvisualized in this chart.

5.2.2 Key resultsThe assessed bioenergy scenarios did not provide substantial GHG emissionssavings compared with the reference (use of the land for spring barley and of

fossil fuels for energy production) when iLUC impacts were accounted for(Figure 5). The iLUC was the major impact on global warming, representing a paramount average of 41% of the induced GHG emissions (about 310 ±170 tCO2-eq. ha-1). This equalled 70-130 g CO2-eq. MJ-1 of solid biofuel produced.This was calculated dividing the GHG emissions by the energy yielded from 1hectare of land and introduced into CHP units (taken relative to the LHV dry basis and assuming the average iLUC value 310 for the calculation). Thesevalues are in the range of the results for liquid biofuels provided by other studiesfocusing on iLUC. For example, Edwards et al. (2010) reported iLUC between

16 and 222 g CO2-eq. MJ-1 for RME and bioethanol (cultivated in EU and US,see also Table 7 in section 6).

Only co-firing of willow and Miscanthus achieved GHG savings compared withthe reference. However, these were far from the 35% GHG reduction targetenforced by European Union (2009) and here used as a hypothetical reference forthe purpose of comparison. These GHG results are lower than those reported inother studies on bioenergy. For example, a number of reviewed bioenergy studies(Brandao et al., 2010, Fazio and Monti, 2011, Styles and Jones, 2007) found



40

GHG savings between ca. -10 and -35 t CO2-eq. ha-1 y-1 (i.e. ca. between -200 and-700 t CO2-eq. ha-1 in 20y). The reason lies on the fact that these studies did notinclude iLUC impacts thereby dramatically underestimating the actual GHG

emissions.

In the eutrophication categories (N and P related) the performance was dictated by two main forces: i) dLUC impacts (difference between cultivating the energycrop and spring barley) and ii) leaching of nutrients from use on land of thedigestate (anaerobic co-digestion scenarios). With respect to aquatic N-eutrophication, fewer requirements for N determined savings in the bioenergyscenarios involving thermal conversion of willow and Miscanthus . Lowersavings were seen in the anaerobic digestion scenarios due to N leaching.Conversely, all the bioenergy scenarios based on ryegrass showed net impacts asa consequence of higher N-fertilizers requirement compared with the reference.

With respect to aquatic P-eutrophication, only the bioenergy scenarios involvinganaerobic digestion showed net impacts as a consequence of P leaching. Thereason for this was that P was applied in excess compared with the average cropsneeds. In fact, the higher nutrients content of the produced digestate involved thatrelatively more P was applied in excess in the co-digestion scenarios comparedwith the reference (use on land of raw pig manure), thus decreasing the overall P-saving potential (Figure 2d) and increasing leaching instead (Figure 2c). Thesavings on the P as resource category were relevant only for ryegrass due to thefact that no P-fertilizers are required during the cultivation stage as opposed tospring barley and to the other perennials considered.

The recommendation that can be drawn from this study is to limit energy cropscultivation and prioritize instead the use of biomasses which do not involveupstream impacts associated with changes in the use of the land, such as organic

and garden waste, manure, wood residues, straw and other agricultural residues.Further, if energy crops are to be used, then highly-efficient co-firing of willowand Miscanthus are the most favourable strategies for Danish conditions.



41

Figure 6. LCA results for a) global warming (100y horizon, t CO2-eq. ha-1); b) aquatic N-eutrophication (kg N ha-1); c) aquatic P-eutrophication (kg P ha-1); and d) phosphorous asresource (kg P ha-1). All systems represent a 20 year time scope. From Tonini et al. (II ).



42

5.3 Case study: energy from MSWIn the light of the significant environmental impacts associated with energy crops production (Tonini et al., II ), optimizing energy production from waste (e.g.MSW) becomes important in the endeavour of decreasing fossil fuelconsumption and resource depletion. The environmental impacts associated witha range of technologies for energy production from MSW were investigated inTonini and Astrup (III ), Montejo et al. (V) and Tonini et al. (VI ). These includedrelevant waste treatment technologies such as incineration, conventionallandfilling, landfilling in bioreactor and mechanical-biological treatment. InTonini et al. (IV ) the waste refinery process was investigated by means of acombination of sampling, waste characterization and MFA/SFA/EFA.

5.3.1 Modeling aspectsIn Tonini and Astrup (III ) and Tonini et al. (IV ) and (VI ) the focus of theassessment was the waste refinery process. Table 5 provides an overview of thewaste management scenarios used as reference in the LCAs and compared withthe waste refinery scenarios. For the latter, a number of alternatives wereconsidered: this included co-firing (alternatively called co-combustion) of the bioliquid, anaerobic digestion of the bioliquid with further use of the produced biogas in gas engine or in vehicles, co-firing and incineration of the residualsolid. Sorting, recovery and further recycling of selected materials from the solidfraction prior to combustion was also considered.

In Tonini et al. (IV ) a waste sampling campaign followed by wastecharacterization (waste and chemical composition) was performed. Thisconsisted on 4 days of sampling at a pilot-scale waste refinery. After, wastecharacterization was performed: a number of waste material fractions were handsorted from the daily collected sample and the relative weight measured. The

selected waste material fractions were then dried and grinded. Further mixing andmass fractional reduction was performed until the mass required for chemicalanalysis was obtained. Selected chemicals were then analyzed; these includedfossil carbon content (represented by the14C content in 12C) analyzed byaccelerated mass spectrometry (AMS). The data were further elaborated byimplementing an original mathematical model facilitated by Matlab 2010. Thiswas needed in order to quantify the potential for bioliquid and associateddownstream biogas recovery as a result of a post-treatment of the ‘ex-enzymatictreatment’ outputs. This could consist of, for example, washing, sieving and



43

pressing. The quality of the bioliquid and its residue after anaerobic digestion(i.e. digestate) were of central focus in this study.

Table 5 . Overview of the method and reference scenarios used to assess the environmental performance of the waste refinery process. The options considered for the energy conversion of bioliquid and residual solid are also reported. CC: co-combustion; AD GE: anaerobic digestionand use of the gas in gas engine for CHP; AD TF: anaerobic digestion and use of the gas fortransport; INC: incineration; FE: ferrous metals; AL: aluminium; HP: hard plastic; GL: glass; NFE: non-ferrous metals. *Only assessed in the sensitivity analysis.

No Method Reference scenario(s)Options for the waste refinery productsBioliquid Residual solid Materials recycled

III LCA, EFA Incineration1. CC2. AD GE

1. CC2. INC

FE, AL, HP, GL

IVMFA, SFA,EFA - AD GE INC FE, NFE, HP

V LCA, MFA Eight MBT scenarios - - -

VI LCA, EFA

1. Incineration2. Conventional landfilling3. Landfilling in bioreactor4. MBT

1. AD GE2. AD TF

INC FE, AL, HP*

In Montejo et al. (V) the focus was instead on the current performances and onthe potential for improvements of existing MBT plants operated in Castilla yLeon (Spain). Eight MBT plants having different operational conditions wereinvestigated. The MBT plants aim at stabilizing the OFMSW prior to landfillingand to produce a RDF fraction. The RDF is currently landfilled (referencescenario). The main difference across the 8 plants was the biological treatment ofthe OFMSW: four of the plants used anaerobic digestion and post-compostingand four performed direct composting. Waste sampling and characterization wasused to define the individual waste compositions and to quantify the overall mass

flows and recovery efficiencies (e.g. for OFMSW, paper, plastic, metals, etc.).The inventoried data were finally used to compare the current plants performances and to suggest process optimizations based on the LCA results.

5.3.2 Key results – focus on the waste refineryAs highlighted in Tonini et al. (IV ), from 1 tonne of dry MSW the waste refinerymay recover ca. 56% of the input dry matter as bioliquid yielding 6.2 GJ biogas-energy, corresponding to about 690 kWh electricity and 3,100 MJ heat producedfrom state-of-the-art gas engines. Additional energy may be recovered fromincineration of the residual solid, which may include selected plastic fractions if



44

not sorted out. The recovery of biogenic carbon and nutrients (N, P and K) in thedigestate left from bioliquid digestion may be between 83% and 93%. Thequality of the digestate may be of concern for the content of selected metals (Cd

and Ni) and other organic compound (e.g. DEHP, i.e. di(2-ethylhexyl)phthalate).These may represent a limitation for use on land.

From a GHG perspective, the results from Tonini et al. (III ) and Tonini et al.(VI ) showed that the waste refinery scenarios may achieve benefits comparableto state-of-the-art incineration and MBT. When coal was the marginal fuel forelectricity production, use of the biogas for CHP showed better performancesthan use for transport. When natural gas was the marginal, the GHG performances were comparable (Tonini et al., (VI )). Tonini et al. (VI ) alsohighlighted that the performance on GW is significantly affected by the wastecomposition. The waste refinery scenarios performed better than incineration provided waste compositions with high organic content (ca. 70%); viceversawhen the organic content was lower. In Tonini et al. (VI ) it was alsodemonstrated that, even in the case of waste composition having high organiccontent, incineration and MBT scenarios may perform comparably to wasterefining scenarios (GHG-wise) provided implementation of organic wastesource-segregation. These findings can be observed in Figure 7.

Electricity production is increased in waste refining scenarios compared withstate-of-the-art incineration. Based on modeling, Tonini et al. (VI ) reported a netincrease of 15-40%, depending upon the organic content. The lower end is betterrepresentative of Danish rMSW. The increase in electricity production comes atthe expenses of the heat generation, which is decreased by 19-30% (Tonini et al.,VI ). Though electricity production is improved compared with incineration, theimpacts associated with use of energy and enzymes for the pre-treatment finallylead to GHG benefits comparable to incineration.

P resource savings through use on land of the digestate (from bioliquid digestion)is possible within the waste refinery scenarios (see Figure 7). These may beoptimized compared with alternatives based on organic waste source-segregation(Tonini et al., VI ). However, the bottleneck here is represented by leaching ofnitrogen (Tonini et al.,III ) and potential contamination of the digestate (Toniniet al., IV ) which impact on nutrient enrichment and toxicity categories. Theimpact on nutrient enrichment may be limited by introducing a post-composting phase for the digestate (Tonini et al.,VI ). This is in accordance with Boldrin etal. (2011). Yet, the concentration of heavy metals and other hazardous chemicals



45

could represent an impediment to use on land. Alternative solutions for thetreatment of the digestate or compost are incineration, landfilling and co-digestion. The first and the second, however, would determine losses of P

(although P extraction from bottom ash may be envisioned). In this perspective,co-digestion of the bioliquid with animal manure may be seen as a potentialsolution, though the overall load of hazardous substances on soil would beunchanged (but the concentrations would likely decrease below selected limitvalues).

Table 6 highlights the environmental impacts potentially induced by wasterefining scenarios in selected environmental categories. The results are fromTonini and Astrup (III ) and Tonini et al. (VI ). For the waste refinery it isassumed that the biogas is used for CHP and the residual solid is incinerated(also CHP); ferrous and non-ferrous metals are assumed sorted and recycled. A brief explanation of why waste refineries may perform better is also given.

Table 6 . Overview of the performance of the waste refinery scenarios compared withincineration (INC), mechanical-biological treatment (MBT) and landfilling in bioreactor (BLF).↓: decreased impacts;↑: increased impacts;↕: increased/decreased depending upon selectedLCA parameters; GW: global warming; AC: acidification; NE: nutrient enrichment; ETwc:ecotoxicity in water, chronic; HTw: human toxicity via water; HTs: human toxicity via soil;Pres: P resource savings; ST: stored toxicity; GWres: groundwater resource.

Category vs. INC vs. MBT vs. BLF Reason for WR being better:GW ↕ ↓ ↓ Improved electricity recovery

AC ↕ ↓ ↓ Improved electricity recovery

NE ↑ ↑ ↕ N leaching through use on land

ETwc ↓ ↕ ↓ Improved Al and energy recovery

HTw ↑ ↑ ↑ Metals applied on soil through use on land

HTs ↑ ↑ ↑ Metals applied on soil through use on land

Pres ↓ ↓ ↓ P savings through use on land

ST ↕ ↓ ↓ Avoided storage of metals in landfill

GWres ↓ ↓ ↓ Avoided leaching from landfill



46

Figure 7 . Non-toxic LCA results from Tonini et al. (VI ) (coal as marginal for electricity andnatural gas as marginal for heat). The waste refining scenarios (in bold) are compared with:INC: incineration, CLF: conventional landfilling, BLF: landfilling in bioreactor, MBT AC:MBT with anaerobic digestion of the OFMSW, MBT DC: MBT with direct composting of theOFMSW, WR GE: waste refining with use of the biogas in gas engine, WR TF: waste refiningwith upgrading of the biogas to transport fuel and use in vehicle. (0): no organic source-segregation; (II): 70% efficiency of organic source-segregation (plus 20% mass loss duringmechanical pre-treatment) for further anaerobic digestion and use on land; (III): 100% organicsource-segregation (0% pre-treatment loss) for further anaerobic digestion and use on land; GW:global warming. AC: acidification; NE: nutrient enrichment; Pres: P resource saving. (a):Danish waste composition (with low organic content); (b): Spanish waste composition (withhigh organic content). The dashed line indicates the performance of incineration without organicsource-segregation (reference scenario for DK).



47

5.3.3 Key results – focus on MBTThe results from Montejo et al. (V) indicated that the actual operationalconditions (i.e. energy recovery, final use of the biogas, material separation)largely affect the environmental performance of MBT plants: the efficiency ofmaterials and energy recovery in fact varied widely across the scenarios underassessment. Based on actual operational data, there was no clear evidence, forexample, that MBT with anaerobic digestion of the OFMSW performed betterthan MBT with direct composting. On GW the current performance dependedupon two main factors: i) the final use of the (eventually produced) biogas (e.g. ashare may be used for heating the digestion reactor) and ii) the materials recoveryefficiency. It is clear that an optimal energy recovery from the OFMSW may

rank MBT with anaerobic digestion as favourable option over MBT withcomposting, provided equal material recovery. However, the actual operationalmodes may lead to recovery efficiencies far lower than expected (Montejo et al.,(V)).

The largest potential for increasing GHG savings is connected to the optimizationof material and energy recovery. This could be done by: i) improving materialsand energy recovery at MBTs and ii) encouraging source-segregation strategies.With respect to the first, Montejo et al. (V) estimated the maximal (average of all plants investigated) potential for GHG savings from recovery/recycling to be between -170 and -240 kg CO2-eq. t-1 ww depending on whether C sequestrationfrom paper in the landfill was accounted for or not. Optimization of the biological treatment with increased methane yield and improved electricityrecovery also provided additional savings (between -8 and -93 kg CO2-eq. t-1 ww).

The results for possible RDF management strategies were dramatically

influenced by the assumptions on energy system (marginal energy) and carbonsequestration in landfill. These affected the results on GW. Under the assumptionthat natural gas was the marginal for electricity production (along withaccounting the carbon sequestered in landfill), RDF incineration resulted worsethan landfilling from a GHG perspective. This was not the case when coal wasthe marginal: under this assumption RDF incineration performed better. In all theremaining environmental categories RDF incineration resulted in better performances compared with landfilling. If RDF-to-energy is to be implemented,a possible solution may consist in utilizing the RDF in cement kilns to substitute



48

for heavy oil fuel (‘direct’ CO2 emission factor of ca. 78 g CO2 MJ-1): this wouldlead to GHG savings comparable with landfilling.Source-segregation (ii) should also be encouraged, considered the fact that the

current recycling rate in the region is ca. 12% (i.e. rMSW constitutes ca. 88% ofthe generated MSW). Tonini et al. (VI ) modelled a number of scenarios (126) based on the average Castilla y Leon waste composition from Montejo et al.(2011). The results highlighted overall GHG savings of ca. -500/-800 kg CO2-eq.t-1 ww when source-segregation of paper, plastic, metals, glass and organic wastewas implemented for further recycling along with treatment of the rMSW inMBTs provided with efficient energy recovery from the produced biogas andRDF. These GHG savings are considerably increased compared with the currentaverage plants performance (ca. -240 kg CO2-eq. t-1 ww). In addition, source-segregation of OFMSW may dramatically increase the quality of the producedcompost which is currently used as daily cover in landfills (Montejo et al., 2010)due to its low quality (contamination with impurities).

These results highlight one important aspect: in many EU regions MSW is stillregarded as a waste to be safely disposed rather than as a carrier of resource andenergy. Often, technologies such as MBT and incineration, even if recentlycommissioned, aim at a safe disposal (e.g. stabilization of the OFMSW prior tolandfilling) rather than at maximizing energy and materials recovery. In this perspective, large potentials exist to optimize the environmental performance oftreatment technologies and source-segregation strategies, therefore changing the perception on MSW from ‘waste to be disposed’ to ‘resource and energy carrier’.



49

6. Discussion The ambition of Denmark and Europe is to reduce fossil fuel dependency andresource depletion. The road towards more renewable energy systems is howeverchallenged by numerous aspects associated with primary fuel supply anddiminished energy consumption (demand). If solid and liquid biofuels are neededwithin the energy system, cultivation of energy crops is likely to be required.This may induce additional environmental impacts and/or shift the burdens fromone environmental compartment to another. Within this thesis, LCA modelingwas used to evaluate potential benefits and drawbacks of future energy systems,energy crops and innovative WtE technologies.

The most critical aspect in LCA of bioenergy systems is associated with thequantification of LUC. Particularly, the estimation of iLUC impacts is associatedwith high uncertainty and to date there is no agreement on a recommendedmethodology. In Tonini and Astrup (I) the iLUC impacts were quantified afterSchmidt (2008) and Schmidt (2007), and estimated to about 220 t CO2-eq. ha-1.This corresponded to ca. 120 g CO2-eq.-MJ-1 FT-biodiesel from willow and 230 gCO2-eq.-MJ-1 RME-biodiesel. The approach used in Tonini et al. (II ) was instead based on the results for Denmark of a general equilibrium model (GTAP)(Kloeverpris, 2008). The corresponding iLUC impact was ca. 310 ±170 t CO2-eq.ha-1 (ca. 70-130 g CO2-eq.-MJ-1 depending upon crop and conversion pathway).The fundamental difference between the two approaches lies on assumptionsregarding the market elasticity (i.e. how much Danish barley is replaced or, inother words, the substitution ratio) and on the method used to identify the landconverted: Schmidt (2008) uses a 1:1 substitution ratio (market elasticity equal to1) and suggests Canada (and particularly Canadian grassland) as marginal barleysupplier based on the predictions of FAPRI (2006). Instead, a general equilibriummodel is used in (Kloeverpris, 2008) to identify amount and location of landconverted as well as related biomes. Partial/general equilibrium modelsinherently assume a market elasticity < 1 (and therefore a substitution ratio < 1).Since the mechanisms involved in land use changes are complex, the use ofgeneral or partial equilibrium models as supporting tools is generally arecommended approach (Edwards et al., 2008, Edwards et al., 2010).



50

Table 7 . Life cycle and associated iLUC GHG emissions (g CO2-eq. MJ-1 biofuel input tovehicles or CHP units) for different bioenergy systems considered in this thesis. For comparison purposes, the GHG emissions of selected fossil fuel systems are also reported. [1]: Tonini andAstrup (I); [2]: Tonini et al. (II ); [3]: Fritsche (2008); [4]: Edwards et al. (2010); n.r.: notreported. Note that values from Tonini and Astrup (I) and Tonini et al. (II ) are rounded.

[1], [2] [3] [4]

Life cycle* iLUC Life cycle* iLUC Life cycle* iLUC

Liquid biofuels and fossil fuels for transport: production and combustion

RME from rapeseed 290 230ε 73-168δ 34-102δ n.r. 57-222ζ

FT-biodiesel (willow) 65-88α 120ε 36-145δ 39-117δ n.r. -

Ethanol from wheat n.r. n.r. 79-173δ 34-102δ n.r. 16-155ζ

Diesel 85-90η (Life cycle emission)

Gasoline 87-90η (Life cycle emission)

Solid biofuels for electricity and heat generation: production and combustion

Ryegrass to CHP 210(180-250)β

110(91-130)γ n.r. n.r. n.r. n.r.

Willow to CHP 140(120-170)β

85(70-100)γ n.r. n.r. n.r. n.r.

Miscanthus to CHP 160(140-190)β

110(91-130)γ n.r. n.r. n.r. n.r.

Coal to electricity CP 110η

(Life cycle emission) Natural gas-heat boiler 75η (Life cycle emission)

*Total life cycle GHG emissions (including iLUC, dLUC and use of byproducts).α With and without considering benefits from biochar.β, γ Depending on the energy conversion process (e.g. anaerobic digestion, combustion, co-firing, etc.). The lower and upper bound corresponds here to co-firing and anaerobic digestion, respectively.δ Theauthors distinguish between low iLUC (25% of the non-zero risk biofuels are subject to a ‘full iLUC factor’ of 13.5 tCO2-eq. ha-1 y-1) and high iLUC (75% of the non-zero risk biofuels are subject to the ‘full iLUC factor’). The rangeof values reported is for cultivation on arable and grassland in EU.ζ overall range of iLUC results obtained fromGTAP (upper bound) and FAPRI-CARD (lower bound).η From the Ecoinvent database v2.2.ε Note that it iscalculated dividing the iLUC by the energy content of the biodiesel produced.

An overview of the iLUC GHG emissions reported by a number of studies(including this thesis) is presented in Table 7. These, despite the significantuncertainties and differences associated with the iLUC approaches, all highlightiLUC as the most important contributor to the induced GHG emissions in bioenergy systems. The values from Fritsche (2008) distinguish between low andhigh iLUC. Note that, for RME, the energy yield considered in Fritsche (2008) ishigher than in Tonini and Astrup (I) (ca. 100 vs. 50 GJ ha-1). This determinessignificantly lower iLUC and life cycle GHG impacts in Fritsche (2008).



51

Note that GHG emissions for liquid biofuels were calculated relative to theenergy content of the biodiesel produced, while for solid biofuels the GHGemissions were calculated relative to the energy content of the dry biomass

produced by 1 ha and input to CHP. Therefore, values of iLUC and life cycleGHG emissions of liquid and solid biofuels in Table 7 should not be directlycompared.

Tonini and Astrup (I) assessed a number of future Danish energy scenarios basedon high shares of biomass and wind energy. The scenarios resulted from ESAassuming a number of technical measures to reduce the final energy demand (e.g.decommissioning of old power plants and use of fuel cells, improved insulationof buildings, improved efficiency of transport means, reduced district heatinglosses, etc.). The analysis focused on the interactions between the energy systemand other sectors of the economy/society. For example, if grass is to be used forenergy, the consequence of this would translate into some other crop to be usedfor feeding and bedding if this is the current management of the grass underassessment; the final effect may be land conversion to produce animal feed (seesection 3). The same approach may apply to the case of straw. Such effects wereconsidered into the assessment.

Overall, a combination of reduced energy demand and replacement of fossil fuelwith wind energy and domestically available biomass may reduce by more thanhalf the GHG emissions compared with current levels. However, this may shiftthe burden from global warming to eutrophication effects. In addition, the mainchallenge of future energy system is related to provision of diesel-like biofuelsfor transport. RME-biodiesel was found to induce higher GHG emissions (ca.290 g CO2-eq. MJ-1 fuel) than fossil diesel (ca. 89 g CO2-eq. MJ-1 fuel) alongwith increased eutrophication and acidification effects. FT-biodiesel fromlignocellulosic crops may also cause GHG emissions comparable to those of

fossil diesel. In the light of these findings, energy production from waste andresidual biomasses (not involving iLUC) should be prioritized and, to the extentneeded, a share of the transport sector should still rely on fossil fuel rather thanon biofuels implying iLUC impacts.

Cultivation of perennial energy crops for combined heat and power productionshould be considered carefully. Tonini et al. (II ) highlighted willow and

Miscanthus co-firing as promising strategies. Ryegrass and anaerobic digestionwere the worst crop and energy-pathway, determining the highest globalwarming and eutrophication impacts. If biogas is for some specific reason



52

required into the energy system (e.g. for fuel cells or transportation) then co-digestion with manure should be encouraged to boost manure-biogas and avoidthe current impacts associated with raw manure management (storage and use on

land). However, the results generally showed that the final GHG performance istremendously affected by the magnitude of the iLUC impact. This is the keyfactor determining the GHG performance of bioenergy systems and furtherresearch into the methodology for quantifying iLUC impacts appears inescapablyneeded.

As opposite to the crops, waste does not induce iLUC impacts. Optimization ofenergy production and material/resource recovery from waste (e.g. MSW) becomes therefore very important. Source-segregation of recyclables (plastic,metals, paper and cardboard, glass and eventually organic waste) may contributeto save on GHG emissions from virgin production. However, even whenapplying these measures, a large share of the generated MSW ends up in therMSW stream (see for example Tonini et al.,VI ). Optimizing energy andmaterial/resource recovery from this stream is therefore essential. This acquiresfurther relevance in the light of the EU directives on waste and landfills (TheEuropean Parliament and The Council, 2008, CEC, 1999) aiming at maximizingmaterial, resource and energy recovery from waste streams and minimizingdisposal of biodegradable waste in landfill. In this perspective innovative pre-treatment technologies such as waste refineries represent promising solutions.

The waste refinery process was investigated during this research by applyingdifferent methods: waste sampling and characterization, LCA, MFA, SFA andEFA. From a GHG perspective the performance of waste refining scenarios wascomparable to incineration and state-of-the-art MBT plants (Tonini and Astrup,III and Tonini et al., VI ). The main advantages of such technology arerepresented by: 1) higher electricity production. 2) Opportunity of optimizing

nutrients recovery (particularly P) compared with other practices: for example,source segregation may induce high nutrients losses (Bernstad, 2012), compostfrom MBT is generally landfilled for its low quality (Montejo et al., 2010) andincineration determines loss of nutrients. 3) Flexibility and storability of the product-fuels (biogas-bioliquid and residual solid).

(1) Biogas production (potential ca. 440 Nm3 CH4 t-1 VS) and associatedelectricity production is increased as an effect of the enzymatic treatmentundergone by the degradable materials. In fact, the typical methane potential fornon-treated paper materials is much lower (ca. 120-250 Nm3 CH4 t-1 VS).



53

However, the increased in gross electricity production (ca. 15% for Danishhousehold waste, see section 5) is compensated by the emissions associated with pre-treatment (energy and enzymes consumption) and the reduction in heat

production, finally leading to overall GHG savings comparable to incineration.

(2) Although recovery of nutrients (particularly P) can be optimized throughwaste refining, application on land of the digestate may be limited by the contentof selected metals (e.g. Cd, Ni and, to a minor extent, Hg and As) and phthalates.Tonini et al. (IV ) highlighted that the concentration of selected hazardouschemicals may be close or exceed the Danish limits for digestate use on land(Danish Ministry of Food, Agriculture and Fisheries, 2006). This also inducessignificant impacts on the LCA toxic categories compared with incineration andother management practices such as landfilling and MBT (although the toxicimpacts from landfills can be seen in other environmental categories, e.g. ‘storedtoxicity’ in the landfill body or long term emissions, see for example Manfredi etal., 2009). A possible solution to reduce the concentration of the contaminants isrepresented by co-digestion with manure though the overall load on soil would be unchanged (i.e. no changes on the LCA impacts). Alternative solutions may be incineration and landfilling, though these would likely cause losses ofnutrients.

(3) Within this thesis, the potential benefits associated with flexibility andstorability of the product-fuels from the waste refinery were not addressed. Thelevel of flexibility of the energy conversion technologies (ramp-up ability,storability of the fuel, ability to switch between outputs) may be significant whenenergy systems with high penetration of fluctuating energy sources (e.g. windenergy) are considered. This requires further investigations.



54



55

7. Conclusion and recommendations This thesis assessed the environmental performance of bioenergy and waste-to-energy system using life cycle assessment modeling. The main findings of theresearch can be summarised as follows:

(1) The amount of energy that can be produced by the available waste and biomass resources is limited compared with society energy needs. Only acombination of reduced energy demand, electrification of the transport sector andreplacement of fossil fuels with biomass and other renewables (primarily windenergy) has the potential to significantly reduce GHG emissions associated withthe energy system.

(2) The production of liquid biofuels for the transport sector represents a bottleneck. Significant GHG and eutrophication potential impacts are associatedwith the production of liquid biofuels from energy crops such as rapeseed orwillow cultivated at the expenses of food crops. When indirect land use changesimpacts are considered, the associated GHG emissions may be comparable oreven exceed the current GHG emissions from gasoline and diesel. In the case ofcombined heat and power production from perennial energy crops the overall lifecycle GHG emissions may be lower than those for liquid biofuels, due toimproved energy conversion efficiency. The iLUC impacts are here crucialdetermining the beneficial or detrimental GHG performance of the bioenergysystems under assessment. In addition, potential eutrophication impacts may be aresult of selected energy conversion pathways involving the return of nutrients tothe land (i.e. digestate from anaerobic digestion).

(3) Existing state-of-the-art and emerging waste-to-energy technologies, such asincineration, MBT and waste refining (with associated downstream energyrecovery processes) may achieve comparable GHG emission savings. Comparedwith incineration, waste refining and MBT improve materials recovery fromwaste streams. P recovery is also possible in waste refineries, although thequality of the digestate/compost obtained from anaerobic digestion of the produced bioliquid may not comply with selected land application limits. For thecase of MBT, use on land of the digestate/compost is typically not allowed as the presence of contaminants (metals) is high. This determines loss of P andnutrients. From a mere energy perspective, waste refineries allow for anincreased electricity production compared with incineration and MBT. Related

GHG savings are however balanced by additional GHG emissions associated



56

with enzymes use and decreased heat production. Moreover, digestate use onland results in additional N-eutrophication effects due to NO3- leaching. This,however, could be significantly reduced if a composting phase is performed prior

to use on land. From an energy-system perspective, the possible advantage ofwaste refining and MBT is the production of storable product-fuels which mayincrease the flexibility of the technology

Based on the conclusion, the following recommendations are provided :

(1) Cultivation and production of liquid (transport) biofuels from energy crops(even perennials) should be limited when inducing indirect land use changes.Production of solid biofuels from selected perennials (e.g. willow and

Miscanthus ) may be environmentally more sustainable. However, these aredramatically dependent upon methodology and assumptions concerning iLUCand the overall GHG benefits may be finally low. With respect to this, furtherresearch is needed.

(2) For municipal solid waste, state of the art incineration, MBT and wasterefining (with associated energy and materials recovery processes) may all provide significant and comparable GHG emission savings. The wastecomposition (e.g. amount of organic and paper) and properties (e.g. LHV, watercontent) may affect the final ranking. When assessing the environmental performance of waste refining it is recommendable to have a detailed knowledgeof the waste composition as this determines the energy outputs and, therefore, theLCA results. Overall, waste refineries maximize electricity production but inducelower heat production and higher pre-treatment costs.

(3) If P recovery is the priority, waste refining is the most suitable technology.However, associated use on land may increase N-eutrophication andcontamination of soil compared with incineration. To this respect, a post-composting phase for the digestate is recommended to significantly reduce NO3

- leaching. Efficient source-segregation of the organic waste can potentially avoidcontamination with metals and recover phosphorous. Recent studies however,show concerns about this strategy as they found that the segregation process maylead to significant mass losses as well as to contamination with impurities.



57

8. PerspectivesThe findings of this thesis provide the basis for further investigations on thefollowing topics:

Flexibility of energy conversion technologiesWaste incinerators operate as base load technologies in the energy system as the possibility of storing the waste over long periods is limited. This may representan important limitation in the perspective of future energy systems with increased penetration of fluctuating energy sources such as wind energy. In this context,flexible energy technologies able to: i) switch between output-products(regulation ability), ii) store energy carriers (storability) and iii) quickly regulate

the production (ramp up ability) may be preferred. In current LCA studies(including this thesis) the potential environmental benefits associated with‘increased flexibility’ are not addressed.

Waste characterization Waste composition (including chemical composition, water content, LHV, etc.)affects the LCA results for waste refineries as the quality and quantity of the bioliquid produced is a function of the content of organic, paper, cardboard andof the presence of metals in the waste input. A detailed knowledge of the wastecomposition provides robustness to the LCA results. Waste and chemicalcomposition data used in LCA need constant update and revision throughexperimental investigations (i.e. sampling and analyses). These should focus onwaste material fractions composition as well as on water content of the waste,calorific value, content of fossil carbon and metals. Notice that the metal contentalso affects the LCA toxic impacts associated with incineration and co-firing plants.

Digestate use on landFurther investigations are needed to evaluate the quality and effects of use onland of the digestate produced from bioliquid digestion. These should focus onthe content of metals and other hazardous substances (e.g. DEHP) and on theleaching of nitrogen from different products such as raw digestate, dewatereddigestate and post-composted digestate. Specific environmental assessmentsshould be performed to evaluate different techniques and strategies to handle thedigestate in the perspective of recovering the nutrients.



58

Optimal use of available biomass resources In the perspective of future energy systems with increased penetration of windenergy, optimal energy conversion pathways for each available biomass type

should be investigated. Focus should be on optimizing energy production andflexibility.

Indirect land use changesThe methodology for the quantification of iLUC should be improved and a broadly accepted ‘recipe’ should be drafted. Based on the results of this research,the iLUC impacts are the key factor determining the final GHG performance of bioenergy systems based on energy crops cultivation.



59

9. References

Ahrenfeldt, J., Henriksen, U., Jensen, T.K., Goebel, B., Wiese, L., Kather, A., Egsgaard, H.,2006. Validation of a continuous combined heat and power (CHP) operation of a two-stage biomass gasifier. Energy Fuels 20, 2672-2680.

Alvarez, J.A., Otero, L., Lema, J.M., 2010. A methodology for optimising feed composition foranaerobic co-digestion of agro-industrial wastes. Bioresource Technol. 101, 1153-1158.

Andersen, J.K., Boldrin, A., Christensen, T.H., Scheutz, C., 2010. Mass balances and life-cycleinventory for a garden waste windrow composting plant (Aarhus, Denmark). WasteManage. Res. 28, 1010-1020.

Arena, U., Di Gregorio, F., Amorese, C., Mastellone, M.L., 2011. A techno-economiccomparison of fluidized bed gasification of two mixed plastic wastes. Waste Manage.31(7), 1494-1504.

Arena, U., Gregorio, F.D., Santonastasi, M., 2010. A Techno-Economic Comparison BetweenTwo Design Configurations for a Small Scale, Biomass-to-Energy Gasification BasedSystem. Chem. Eng. Journal 162, 590.

Arena, U., Mastellone, M.L., Brunner, P.H., 2009. Scenarios of waste management for a wasteemergency area: A substance flow analysis. J. Ind. Ecol. 13, 735-757.

Arena, U., Zaccariello, L., Mastellone, M.L., 2010. Fluidized bed gasification of waste-derivedfuels. Waste Manage. 30, 1212-1219.

Astrup, T., Bilitewski, B., 2010. Pyrolysis and Gasification, in: Christensen T.H. (Ed.), SolidWaste Technology & Management. John Wiley & Sons, Ltd, pp. 502-512.

Astrup, T., Riber, C., Pedersen, A.J., 2011. Incinerator performance: effects of changes in wasteinput and furnace operation on air emissions and residues. Waste Manage. Res. 29, 57-68.

Austen, G.T., 1984. Shreve's Chemical Process Industries. Mc-Graw Hill.

Bentsen, N.S., Thorsen, B.J., Felby, C., 2009. Energy, feed and land-use balances of refiningwinter wheat to ethanol. Biofuel Bioprod. Bior. 3, 521-533.

Bernstad, A., 2012. Household food waste management. PhD thesis. Water and Environmentalengineering Department, Lund University, Lund, Sweden.

Bernstad, A., Malmqvist, L., Truedsson, C., la C. Jansen, J., 2012. Need for improvements in physical pre-treatment of source-separated household food waste. Waste Manage. DOI:

10.1016/j.wasman.2012.06.012.Bessou, C., Ferchaud, F., Gabrielle, B., Mary, B., 2011. Biofuels, greenhouse gases and climate

change. A review. Agron. Sustain. Dev. 31, 1-79.

Bilitewski, B., Oros, C., Christensen, T.H., 2010. Mechanical Biological Treatment, in:Christensen T.H. (Ed.), Solid Waste Technology & Management. John Wiley & Sons, Ltd, pp. 628-638.

Birgisdottir, H., Bhander, G., Hauschild, M.Z., Christensen, T.H., 2007. Life cycle assessmentof disposal of residues from municipal solid waste incineration: Recycling of bottom ash inroad construction or landfilling in Denmark evaluated in the ROAD-RES model. WasteManage. 27, S75-S84.



60

Boerjesson, P., Berglund, M., 2006. Environmental systems analysis of biogas systems - Part I:Fuel-cycle emissions. Biomass Bioenerg. 30, 469-485.

Boldrin, A., Neidel, T.L., Damgaard, A., Bhander, G.S., Møller, J., Christensen, T.H., 2011.Modelling of environmental impacts from biological treatment of organic municipal wastein EASEWASTE. Waste Manage. 31, 619-630.

Brandao, M., i Canals, L.M., Clift, R., 2011. Soil organic carbon changes in the cultivation ofenergy crops: Implications for GHG balances and soil quality for use in LCA. BiomassBioenerg. 35(6), 2323-2336.

Bruni, E., Jensen, A.P., Angelidaki, I., 2010. Comparative study of mechanical, hydrothermal,chemical and enzymatic treatments of digested biofibers to improve biogas production.Bioresource Technol. 101, 8713-8717.

Brunner, P.H., 2012. Substance Flow Analysis: A Key Tool for Effective ResourceManagement. J. Ind. Ecol. 16(3), 293-295.

Brunner PH, Rechberger H., 2004 Practical handbook of material flow analysis. CRC/Lewis,Boca Raton, Florida, US.

CEC, 1999. Council Directive 1999/31/EC on the landfill of waste. Official Journal of theEuropean Communities L 182/1. Available at:http://www.central2013.eu/fileadmin/user_upload/Downloads/Document_Centre/OP_Resources/Landfill_Directive_1999_31_EC.pdf (accessed October 30, 2012).

Cencic, O., Rechberger, H., 2008. Material Flow Analysis with Software STAN. J. Environ.Manage. 18(1), 3-7.

Chen, Z., Ngo, H.H., Guo, W., 2012. A critical review on sustainability assessment of recycledwater schemes. Sci. Total Environ. 426, 13-31.

Cherubini, F., Bird, N.D., Cowie, A., Jungmeier, G., Schlamadinger, B., Woess-Gallasch, S.,2009. Energy- and greenhouse gas-based LCA of biofuel and bioenergy systems: Keyissues, ranges and recommendations. Res. Cons. Recycl. 53, 434-447.

Cherubini, F., Strømman, A.H., 2011. Life cycle assessment of bioenergy systems: State of theart and future challenges. Bioresource Technol. 102, 437-451.

Christensen, T.H., Bilitewski, B., 2010. Mechanical Treatment: Material Recovery Facilities, in:Christensen T.H. (Ed.), Solid Waste Technology & Management. John Wiley & Sons, Ltd, pp. 349-362.

Christensen, T.H., Manfredi, S., Knox, K., 2010. Landfilling: Reactor Landfills, in: ChristensenT.H. (Ed.), Solid Waste Technology & Management. John Wiley & Sons, Ltd, pp. 772-787.

Christensen, T.H., Matsufuji, Y., 2010. Source Segregation and Collection of Source-Segregated Waste, in: Christensen T.H. (Ed.), Solid Waste Technology & Management.John Wiley & Sons, Ltd, pp. 296-310.

Clavreul, J., Guyonnet, D., Christensen, T.H., Quantifying uncertainty in LCA-modelling ofwaste management systems. Waste Manage. 32(12), 2482-2495.

Clift, R., Doig, A., Finnveden, G., 2000. The application of Life Cycle Assessment to IntegratedSolid Waste Management - Part 1 - Methodology. Process Saf. Environ. 78, 279-287.

Dalgaard, R., Schmidt, J.,H., Halberg, N., Christensen, P., Thrane, M., Pengue, W.A., 2008.

LCA of Soybean Meal. Int. J. Life Cycle Assess. 13, 240-254.



61

Damgaard, A., Riber, C., Fruergaard, T., Hulgaard, T., Christensen, T.H., 2010. Life-cycle-assessment of the historical development of air pollution control and energy recovery inwaste incineration. Waste Manage. 30, 1244-1250.

DEA, 2010a. Energistatistik 2009. Energistyrelsen, Danish Energy Agency (DEA),Copenhagen, Denmark. Available at:http://large.stanford.edu/courses/2010/ph240/smillie1/docs/energy-dk-2009.pdf (accessed November 30, 2012).

DEA, 2010b. Forudsaetninger for samfundsoekonomiske analyser paa energiomraade. DanishEnergy Agency (DEA), Copenhagen, Denmark. Available at: http://www.ens.dk/da-DK/Info/TalOgKort/Fremskrivninger/beregningsforudsatninger/Documents/ForFor%C3%A6tninger%20for%20samfunds%C3%B8konomiske%20analyser%20p%C3%A5%20energiomr%C3%A5det%202011.pdf (accessed November 30, 2012).

DEA, 2009. Energistatistik 2008. Energistyrelsen, Danish Energy Agency (DEA), Copenhagen,Denmark. Available at:http://www.avisredaktionen.dk/fileadmin/Pressesite/TwinHeat/Energistatistik_2200.pdf(accessed November 30, 2012).

Danish Ministry of Climate, Energy and Buildings, 2011. Our future energy. Danish Ministry ofClimate, Energy and Buildings, Copenhagen, Denmark.http://www.ens.dk/Documents/Netboghandel%20-%20publikationer/2011/our_future_energy_%20web.pdf (accessed November 30, 2012).

Danish Ministry of Food, Agriculture and Fisheries, 2006. Bekendtgørelse om anvendelse afaffald til jordbrugsformål (Slambekendtgørelsen). Danish Ministry of Food, Agricultureand Fisheries, Copenhagen, Denmark. Available at:https://www.retsinformation.dk/forms/R0710.aspx?id=13056 (accessed October 30, 2012).

Dauber, J., Jones, M.C., Stout, J.B., 2010. The impact of biomass crop cultivation on temperate biodiversity. GCB Bioenergy 2, 289-309.

DONG, 2009. Avedoerevaerket - Groent regnskab 2008. DONG energy, Hvidovre, Denmark.Available at:http://www.dongenergy.com/SiteCollectionDocuments/NEW%20Corporate/PDP/Groenne _regnskaber_2008/Groent_regnskab_2008_Avedoerevaerket.pdf (accessed November 30,2012).

Edwards, R., Mulligan, D., Marelli, L., 2010. Indirect land use change from increased biofuelsdemand. Comparison of models and results for marinal biofuels production from differentfeedstocks. European Commission Joint Research Centre, Ispra, Italy. Available at:http://ec.europa.eu/energy/renewables/studies/doc/land_use_change/study_4_iluc_modellin

g_comparison.pdf (accessed November 30, 2012); DOI 10.2788/54137.Edwards, R., Szekeres, S., Neuwahl, F., Mahieu, V., 2008. Biofuels in the European context:

facts and uncertainties. European Commission Joint Research Centre, Ispra, Italy.Available at: http://ec.europa.eu/dgs/jrc/downloads/jrc_biofuels_report.pdf (accessed November 30, 2012).

EEA, 2011. Opinion of the EEA Scientific Committee on Greenhouse Gas Accounting inRelation to Bioenergy. European environment Agency (EEA).http://www.eea.europa.eu/about-us/governance/scientific-committee/sc-opinions/opinions-on-scientific-issues/sc-opinion-on-greenhouse-gas/view (accessed November 30, 2012).



62

Eggleston, S., Buendia, L., Miwa, K., Ngara, T., Tanabe, K., 2006. IPCC Guidelines for National Greenhouse Gas Inventories. Vol. 5 Waste. IPCC National Greenhouse GasInventories Programme. Institute for Global Environmental Strategies, Hayama,Kanagawa, Japan. http://www.ipcc-nggip.iges.or.jp/public/2006gl/ (accessed November 30,2012).

Ekvall, T., Weidema, B.P., 2004. System boundaries and input data in consequential life cycleinventory analysis. Int. J. Life Cycle Assess. 9, 161-171.

Emery, I.R., Mosier, N.S., 2012. The impact of dry matter loss during herbaceous biomassstorage on net greenhouse gas emissions from biofuel production. Biomass Bioenerg. 39,237-246.

Energistyrelsen, 2012. Technology data for energy plants - Generation of Electricity and DistrictHeating, Energy Storage and Energy Carrier Generation and Conversion. Danish EnergyAgency (DEA), Copenhagen, Denmark. Available at:http://www.ens.dk/Documents/Netboghandel%20-%20publikationer/2012/Teknologikatalog_2012.pdf (accessed November 30, 2012).

European Union, 2009. Directive 2009/28/EC of the European Parliament and of the Council of23 April 2009 on the promotion of the use of energy from renewable sources and amendingand subsequently repealing Directives 2001/77/EC and 2003/30/EC. Official journal of theEuropean Union 52, L 140/16-L 140/16 - L 140/62.

Eurostat, 2011. Europe in Figures - Eurostat Yearbook 2011. DOI:10.2785/12017. Availableat:http://epp.eurostat.ec.europa.eu/portal/page/portal/product_details/publication?p_product _code=KS-CD-11-001 (accessed November 30, 2012).

FAPRI, 2006. FAPRI 2006, US and world agricultural outlook. Food and Agriculture ResearchInstitute, Iowa, United States.

Fazio, S., Monti, A., 2011. Life cycle assessment of different bioenergy production systemsincluding perennial and annual crops. Biomass Bioenerg. 35, 4868-4878.

Fichtner, 2004. The Viability of Advanced Thermal Treatment of MSW in the UK. FichtnerConsulting Engineers Ltd. London, United Kingdom: ESTET. Available at:http://www.esauk.org/reports_press_releases/esa_reports/thermal_treatment_report.pdf(accessed November 30, 2012).

Finnveden, G.A., Hauschild, M.Z., Ekvall, T., Guinee J.B., J., Heijungs, R., Hellweg, S.,Koehler, A., Pennington, D., Suh, S., 2009. Recent developments in Life CycleAssessment. J. Environ. Manage. 91, 1-21.

Finnveden, G., Björklund, A., Moberg, Å., Ekvall, T., Moberg, Å., 2007. Environmental and

economic assessment methods for waste management decision-support: possibilities andlimitations. Waste Manage. Res. 25, 263-269.

Fritsche, U.R., 2010. The “iLUC Factor” as a Means to Hedge Risks of GHG Emissions fromIndirect Land-Use Change Associated with Bioenergy Feedstock Provision. Working paper prepared for BMU, Darmstadt (forthcoming), Oeko-Institute. Available at:http://www.oeko.de/oekodoc/1030/2010-082-en.pdf (accessed November 30, 2012).

Fruergaard, T., Astrup, T., 2011. Optimal utilization of waste-to-energy in an LCA perspective.Waste Manage. 31, 572-582.



63

Fruergaard, T., 2010. Environmental sustainable utilization of waste resources for energy production. PhD thesis. Department of Environmental Engineering, Technical Universityof Denmark, Lyngby, Denmark. Available at:http://orbit.dtu.dk/en/publications/id(3c1011e3-ab99-4931-beeb-bcece1cb8962).html(accessed November 30, 2012).

Fruergaard, T., Christensen, T.H., Astrup, T., 2010a. Energy recovery from waste incineration:Assessing the importance of district heating networks. Waste Manage. 30, 1264-1272.

Fruergaard, T., Hyks, J., Astrup, T., 2010b. Life-cycle assessment of selected managementoptions for air pollution control residues from waste incineration. Sci. Total Environ. 408,4672-4680.

Gigler, J.K., van Loon, W.,K.P., Sonneveld, C., 2004. Experiment and modelling of parametersinfluencing natural wind drying of willow chunks. Biomass Bioenerg. 26, 507-514.

GTAP Website; https://www.gtap.agecon.purdue.edu/default.asp#2 (Accessed November 30,2012).

Hamelin, L., Joergensen, U., Petersen, B.M., Olesen, J.E., Wenzel, H., 2012. Modelling theenvironmental consequences of direct land use changes from energy crops in Denmark: aconsequential life cycle inventory. GCB Bioenergy 4(6), 889-907.

Hamelin, L., Wesnaes, M., Wenzel, H., Petersen, B.M., 2011. Environmental Consequences ofFuture Biogas Technologies Based on Separated Slurry. Environ. Sci. Technol. 45, 5869-5877.

Hansen, T.L., Bhander, G.S., Christensen, T.H., Bruun, S., Jensen, L.S., 2006. Life cyclemodelling of environmental impacts of application of processed organic municipal solidwaste on agricultural land (Easewaste). Waste Manage. Res. 24, 153-166.

Hughes, Larson, 1998. Effect of fuel moisture content on biomass-IGCC performance. J. Eng.Gas Turbines Power 120, 455-459.

Huijbregts, M.A.J., Gilijamse, W., Ragas, A.M.J., Reijnders, L., 2003. Evaluating uncertainty inenvironmental life-cycle assessment. A case study comparing two insulation options for aDutch one-family dwelling. Environ. Sci. Technol. 37, 2600-2608.

Hulgaard, T., Vehlow, J., 2010. Incineration: Process and Technology, in: Christensen T.H.(Ed.), Solid Waste Technology & Management. John Wiley & Sons, Ltd, pp. 363-392.

IEA, 2011. CO2 emissions from fuel combustion 2011 - highlights. International EnergyAgency (IEA), Paris, France. Available at:http://www.iea.org/co2highlights/co2highlights.pdf (accessed November 30, 2012).

ISO 14040, 2006a. Environmental Management-Life Cycle Assessment-Principles andFramework. 2006-07-01, 2nd ed.

ISO 14040, 2006b. Environmental Management-Life Cycle Assessment-Requirements andGuidelines. 2006-07-01, 1st ed.

Jansen, J.l.C., 2010. Anaerobic Digestion: Technology, in: Christensen T.H. (Ed.), Solid WasteTechnology & Management. John Wiley & Sons, Ltd, pp. 601-617.

Jensen, J.W., Felby, C., Jørgensen, H., Rønsch, G.Ø., Nørholm, N.D., 2010. Enzymatic processing of municipal solid waste. Waste Manage. 30, 2497-2503.

Jirjis, R., 1995. Storage and drying of wood fuel. Biomass Bioenerg. 9, 181-190.



64

Jørgensen, U., Sørensen, P., Adamsen, A.P., Kristensen, I.T., 2008. Energi fra biomasse -Ressourcer og teknologier vurderet i et regionalt perspektiv. University of Aarhus, Aahrus,Denmark. Available at: http://www.naturstyrelsen.dk/NR/rdonlyres/ED42A888-6B5F-4D13-AF5B-B7852976264F/0/Energifrabiomasse.pdf (accessed November 30, 2012).

Jungbluth, N., Chudacoff, M., Dauriat, A., Dinkel, F., Doka, G., Faist, E.M., Gnansounou, E.,Kljun, N., Schleiss, K., Spielmann, M., Stettler, C., Sutter, J., 2007a. Life Cycle Inventoriesof Bioenergy. Swiss Centre for Life Cycle Inventories, ESU-services, Uster, CH Ecoinventreport No 17.

Jungbluth, N., Frischknecht, R., Emmenegger, M.F., Steiner, R., Tuchschmid, M., 2007b. LifeCycle Assessment of BTL-fuel production: Inventory Analysis. RENEW, Renewable fuelsfor advanced powertrains; ESU-services, Uster, Switzerland. Available at: http://www.esu-services.ch/fileadmin/download/jungbluth-2007-Del_5_2_15-LCA-FinalReport.pdf(accessed November 30, 2012).

Kloeverpris, J.H., 2008. Consequential life cycle inventory modelling of land use induced bycrop consumption. PhD thesis, Technical University of Denmark (DTU), Lyngby,Denmark. Available at:http://orbit.dtu.dk/fedora/objects/orbit:82383/datastreams/file_4935915/content (accessed November 30, 2012).

Kofman, P., Spinelli, R., 1997. Storage and handling of willow from SRC. ELSAMPROJEKT87-986376-2-2.

Larsen, J., Østergaard Petersen, M., Thirup, L., Wen Li, H., Krogh Iversen, F., 2008. The IBUSProcess - Lignocellulosic Bioethanol Close to a Commercial Reality. Chem. Eng. Technol.31, 765-772.

Lewandowski, I., Clifton-Brown, J., Scurlock, J.M.O., Huisman, W., 2000. Miscanthus:European experience with a novel energy crop. Biomass Bioenerg. 19, 209-227.

Lewandowski, I., Heinz, A., 2003. Delayed harvest of miscanthus: influences on biomassquantity and quality and environmental impacts of energy production. Eur. J. Agron. 19,45-63.

Lund, H., Mathiesen, B.V., 2009. Energy system analysis of 100% renewable energy systems:The case of Denmark in years 2030 and 2050. Energy 34, 524-531.

Lund, H., 2007. Renewable energy strategies for sustainable development. Energy 32, 912-919.

Lund, H., 2010. Renewable Energy Systems: The Choice and Modeling of 100% RenewableSolutions. Elsevier, Amsterdam, the Netherlands.

Manfredi, S., Christensen, T.H., 2009. Environmental assessment of solid waste landfillingtechnologies by means of LCA-modeling (EASEWASTE). Waste Manage. 29, 32–43.

Mathiesen, B.V., Lund, H., Hvelplund, F.K., Connolly, D., Bentsen, N.S., Tonini, D.,Morthorst, P.E., Wenzel, H., Astrup, T., Meyer, N.I., Münster, M., Østergaard, P.A., Bak-Jensen, B., Nielsen, M.P., Schaltz, E., Pillai, J.R., Hamelin, L., Felby, C., Heussen, K.,Karnøe, P., Munksgaard, J., Pade, L., Andersen, F.M., Hansen, K., 2011. CEESA 100%Renewable Energy Scenarios towards 2050. Aalborg University. Available from:http://www.ceesa.plan.aau.dk (accessed November 30, 2012).



65

Mathiesen, B.V., Lund, H., Karlsson, K., 2009. IDAs Klimaplan 2050, baggrundsrapport –Tekniske systemanalyser, brændselsforbrug, drivhusgasser, samfund søkonomiskekonsekvenser, erhvervspotentialer, beskæftigelseseffekter samt helbredsomkostninger(IDAs Climate Plan 2050, background report in Danish and English). Danish Society ofEngineers (IDA, Ingeniørforeningen Danmark), Copenhagen, Denmark. Available at:http://ida.dk/omida/laesesalen/Documents/UK-Future%20Climates%20Background%20Report.pdf (accessed November 30, 2012).

Mathiesen, B.V., Lund, H., Karlsson, K., 2011. 100% Renewable energy systems, climatemitigation and economic growth. Appl. Energy 88, 488-501.

Mathiesen, B.V., Münster, M., Fruergaard, T., 2009. Uncertainties related to the identificationof the marginal energy technology in consequential life cycle assessments. J. Clean. Prod.17, 1331-1338.

Mazzoleni, C., Kuhns, H.D., Moosmuller, H., Witt, J., Nussbaum, N.J., Chang, M.C.O.,Parthasarathy, G., Nathagoundenpalayam, S.K.K., Nikolich, G., Watson, J.G., 2007. A casestudy of real-world tailpipe emissions for school buses using a 20% biodiesel blend. Sci.Total Environ. 385, 146-159.

Mcgechan, M.B., 1989. A review of losses arising during conservation of grass forage.1. Fieldlosses. J. Agr. Eng. Res. 44, 1-21.

Moeller, J., Thoegersen, R., Kjeldsen, A.M., Weisbjerg, M.R., Soeegaard, K., Hvelplund, T.,Boersting, C.F., 2000. Foddermiddeltabel Rapport nr. 91. Landbrugets Raadgivningscenter;Landskontoret for Kvaeg, Aahrus, Denmark.

Montejo, C., Ramos, P., Costa, C., Marquez, M.C., 2010. Analysis of the presence of impropermaterials in the composting process performed in ten MBT plants. Bioresource Technol.101, 8267-8272.

Montejo, C., Costa, C., Ramos, P., Marquez, M.C., 2011. Analysis and comparison of municipalsolid waste and reject fraction as fuels for incineration plants. Appl. Therm. Eng. 31, 2135-2140.

Mshandete, A., Kivaisi, A., Rubindamayugi, M., Mattiasson, B., 2004. Anaerobic batch co-digestion of sisal pulp and fish wastes. Bioresource Technol. 95, 19-24.

Müller-Wenk, R., Brandao, M., 2010. Climatic impact of land use in LCA - carbon transfers between vegetation/soil and air. Int. J. Life Cycle Assess. 15, 172-182.

Nges, I.A., Escobar, F., Fu, X., Bjornsson, L., 2012. Benefits of supplementing an industrialwaste anaerobic digester with energy crops for increased biogas production. WasteManage. 32, 53-59.

Nielsen, M., Nielsen, O.-., Thomsen, M., 2010. Emissions from decentralised CHP plants 2007. National Environmental Research Institute, Aarhus University, Aarhus, Denmark.Available at: http://www.dmu.dk/Pub/FR786.pdf (accessed November 30, 2012).

OECD, 2012. Eurostat Joint Questionnaire on waste. Available at: http://glossary.eea.europa.eu//terminology/sitesearch?term=waste 2012 (accessed November 30, 2012).

Oxford Dictionaries, 2012. http://oxforddictionaries.com 2012.

Panoutsou, C., Eleftheriadis, J., Nikolaou, A., 2009. Biomass supply in EU27 from 2010 to2030. Energy Policy 37, 5675-5686.

Pires, A., Martinho, G., Chang, N., 2011. Solid waste management in European countries: Areview of systems analysis techniques. J. Environ. Manage. 92, 1033-1050.



66

Prochnow, A., Heiermann, M., Plochl, M., Linke, B., Idler, C., Amon, T., Hobbs, P.J., 2009.Bioenergy from permanent grassland - A review: 1. Biogas. Bioresource Technol. 100,4931-4944.

Rechberger, H., 2010. Incineration: Co-Combustion, in: Christensen T.H. (Ed.), Solid WasteTechnology & Management. John Wiley & Sons, Ltd, pp. 476-485.

Reimann, D.O., 2009. CEWEP Energy Report II (status 2004-2007). CEWEP, Bamberg,Germany. Available at:http://avfallnorge.web123.no/article_docs/09_04_29%20final%20version%20CEWEP-Report.pdf (accessed November 30, 2012).

Reimann, D.O., 2006. CEWEP Energy Report I (status 2001-2004). CEWEP, Bamberg,Germany. Available at:http://www.cewep.eu/storage/med/media/statements/106_11_07_06_CEWP-Report_Final_Version.pdf (accessed November 30, 2012).

Riber, C., Christensen, T.H., Rodushkin, I., Spliid, H., 2007. Method for fractional solid-wastesampling and chemical analysis. Int. J. Environ. Anal. Chem. 87, 321-335.

Riber, C., Petersen, C., Christensen, T.H., 2009. Chemical composition of material fractions inDanish household waste. Waste Manage. 29, 1251-1257.

Schmidt, J.H., 2007. Life cycle assessment of rapeseed oil and palm oil. PhD thesis, AalborgUniversity, Aalborg, Denmark.

Schmidt, J.H., 2008. System delimitation in agricultural consequential LCA - Outline ofmethodology and illustrative case study of wheat in Denmark. Int. J. Life cycle Assess. 13,350-364.

Searchinger T., Heimlich R., Houghton R.A., Dong F., Elobeid A., Fabiosa J., Tokgoz S., Hayes

D., Yu T.-H., 2008. Use of U.S. croplands for biofuels increases greenhouse gases throughemissions from land-use change. Science 319, 1238-1240.

Searchinger, T.D., 2010. Biofuels and the need for additional carbon. Environmental ResearchLetters 5, 024007-024007.

St-Clair, S.S., Hillier, J., Smith, P., 2008. Estimating the pre-harvest greenhouse gas costs ofenergy crop production. Biomass Bioenerg. 32, 442-452.

Stentiford, E.I., Hobbis, P.G., Barton, J.R., Wang, Z. Banks, C.J. 2010. Evaluating the effect ofautoclaving on the rate of bioprocessing of municipal solid waste. University of Leeds,Leeds, United Kingdom. Available at:http://archive.defra.gov.uk/environment/waste/residual/newtech/documents/trif-leeds-final.pdf.

Styles, D., Jones, M.B., 2007. Energy crops in Ireland: Quantifying the potential life-cyclegreenhouse gas reductions of energy-crop electricity. Biomass Bioenerg. 31, 759-772.

Sun, J., Caton, J.A., Jacobs, T.J., 2010. Oxides of nitrogen emissions from biodiesel-fuelleddiesel engines. Prog. Energ. Combust. 36, 677-695.

Tchobanoglous, G., Kreith, F., 2002. Handbook of Solid Waste Management, Second Edition.McGraw-Hill, New York, USA.



67

The European Parliament and The Council, 2008. Directive 2008/98/EC of the EuropeanParliament and of The Council of 19 November 2008 on waste and repealing certainDirectives. Available at: http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2008:312:0003:0030:en:PDF(accessed November 30, 2012).

Turconi, R., Butera, S., Boldrin, A., Grosso, M., Rigamonti, L., Astrup, T., 2011. Life cycleassessment of waste incineration in Denmark and Italy using two LCA models. WasteManag. Res. 29, 78-90.

Valentine, J., Clifton-Brown, J., Hastings, A., Robson, P., Allison, G., Smith, P., 2012. Food vs.fuel: the use of land for lignocellulosic 'next generation' energy crops that minimizecompetition with primary food production. GCB Bioenergy 4, 1-19.

Wang, W.G., Lyons, D.W., Clark, N.N., Gautam, M., Norton, P.M., 2000. Emissions from nineheavy trucks fueled by diesel and biodiesel blend without engine modification. Environ.Sci. Technol. 34, 933-939.

Weidema, B., 2003. Market information in life cycle assessment. Environmental project 863.Danish Environmental Protection Agency, Copenhagen, Denmark. Available at: http://www.norlca.org/resources/780.pdf (accessed November 30, 2012).

Weidema, B., Frees, N., Nielsen, A.M., 1999. Marginal production technologies for Life CycleInventories. Int. J. Life Cycle Assess. 4, 48-56.

Wihersaari, M., 2005. Evaluation of greenhouse gas emission risks from storage of woodresidue. Biomass Bioenerg. 28, 444-453.

Willumsen, H., Barlaz, M.A., 2010. Landfilling: Gas Production, Extraction and Utilization, in:Christensen T.H. (Ed.), Solid Waste Technology & Management. John Wiley & Sons, Ltd, pp. 841-857.

Zhu, L., Zhang, W., Liu, W., Huang, Z., 2010. Experimental study on particulate and NOxemissions of a diesel engine fueled with ultra low sulfur diesel, RME-diesel blends andPME-diesel blends. Sci. Total Environ. 408, 1050-1058.



68



69

10. Papers

I. Tonini, D., Astrup, T., 2012. Life-cycle assessment of biomass-basedenergy systems: A case study for Denmark. Appl. Energy 99, 234-246.

II. Tonini, D., Hamelin, L., Wenzel, H., Astrup, T. Bioenergy Productionfrom Perennial Energy Crops: a Consequential LCA of 12 BioenergyScenarios including Land Use Changes. Environ. Sci. Technol. 46(24),13521-13530.

III. Tonini, D., Astrup, T., 2012. Life-cycle assessment of a waste refinery process for enzymatic treatment of municipal solid waste. Waste Manage.32, 165-176.

IV. Tonini, D., Dorini, G., Astrup, T. Advanced material, substance andenergy flow analysis of a waste refinery process. Submitted toBioresource Technol.

V. Montejo, C., Tonini, D., Marquez, C.M., Astrup, T. Mechanical-biologicaltreatment: performance and potentials. A LCA of 8 MBT plants includingwaste characterization. Submitted to J. Environ. Manage.

VI. Tonini, D., Martinez, V., Astrup, T. Potential for waste refineries inEurope. To be submitted to Environ. Sci. Technol.



I

LCA of biomass-based energy systems: A case study for Denmark

Tonini, D., Astrup, T

Applied Energy, 99, 234-246



LCA of biomass-based energy systems: A case study for Denmark

Davide Tonini ⇑ , Thomas AstrupDepartment of Environmental Engineering, Technical University of Denmark, Building 115, DK-2800 Kgs. Lyngby, Denmark

a r t i c l e i n f o

Article history:Received 25 October 2011Received in revised form 14 February 2012Accepted 2 March 2012Available online 15 June 2012

Keywords:LCALUCBiomass potentialEnergy system analysisBiodieselEnvironmental impacts

a b s t r a c t

Decrease of fossil fuel consumption in the energy sector is an important step towards more sustainableenergy production. Environmental impacts related to potential future energy systems in Denmark withhighsharesof windand biomass energy wereevaluatedusinglife-cycleassessment (LCA). Basedon theref-erenceyear 2008, energy scenariosfor 2030and 2050were assessed.For 2050three alternatives forsupplyof transport fuels were considered: (1) fossil fuels, (2) rapeseed based biodiesel, and (3) Fischer–Tropschbased biodiesel. Overall, the results showed that greenhouse gas emissions per PJ energy supplied couldbe signicantly reduced (from 68 to 17 Gg CO 2-eq/PJ) by increased use of wind and residual biomassresources as well as by electrifying the transport sector. Energy crops for production of biofuels and theuse of thesebiofuelsfor heavyterrestrial transportationwere responsible for mostenvironmentalimpactsin the 2050 scenarios, in particular upstream impacts from land use changes (LUCs), fertilizer use and NO x

emissionsfrom the transport sector werecritical.Land occupation(includingLUC effects)caused by energycrop production increased to a range of 600–2100 10 6 m 2/PJ depending on the amounts and types of energy crops introduced. Use of fossil diesel in the transport sector appeared to be environmentally pref-erableover biodieselfor acidication, aquatic eutrophicationand landoccupation. Forglobal warming,bio-diesel production via Fischer–Tropsch was comparable with fossil diesel.

2012 Elsevier Ltd. All rights reserved.

1. Introduction

In many countries, considerable efforts have been made to re-duce greenhouse gas (GHG) emissions within the energy sector aspart of the response to climate changes. Within the recent dec-ades, Denmark has managed to control the energy demand whichtoday is similar to that before the oil crisis in the 1970s (864 PJ).In 2008, the share of fossil fuels in the energy system corre-sponded to about 84% (of the primary energy supply). The shareof oil corresponded to 39%. Overall, about 16% of the primary en-ergy supply was based on renewables such as biomass, solar en-ergy and waste resources [1] (for instance, about 20% of theelectricity production was based on wind). The long-term political

target for Denmark is to reach a 100% renewable energy system in2050 primarily based on wind power and biomass energy but alsoinvolving signicant decreases in the national energy demand [2] .Several studies have modeled future sustainable energy systemsfrom a technical perspective [3–12] . According to these studies,100% renewable energy systems can only realistically be achievedthrough signicant reductions in energy demand, increased ef-ciencies of fuel conversion technologies, higher shares of windpower (e.g. up to 50%), replacement of fossil fuels with biomassresource and integration of the transport sector into the energy

system, e.g. through establishment of electric vehicles [13,14] .Although the primary focus of studies involving energy systemanalysis is on the technical design of the energy system (modelingof energy demand and supply, fuels requirements and technologyimplementation), many of these studies also report associated CO 2emissions as an indicator for the environmental impacts related tothe energy system in question. However, such calculations of di-rect emissions associated with the combustion of fuels do not ac-count for important upstream or downstream environmentalimpacts related to the energy system, for example land usechanges (LUCs, due to energy crops cultivation), cascading effects(e.g. substitution of products in the market with byproducts frombiofuel production), and utilization of residues (e.g. digestate and

biochar).GHG emissions have received considerable attention recently;

however, other potential environmental impacts are associatedwith energy production (e.g. eutrophication, acidication and landuse). Such impacts are typically not considered by energy systemanalysis. To provide a full overview of the environmental conse-quences of changing energy production in the future, all upstream,direct and downstream emissions have to be accounted in a life-cycle perspective. We have found no such studies in the literaturefocusing on energy systems with high shares of wind and biomassenergy.

This study quanties the environmental impacts associatedwith potential future energy scenarios for Denmark in 2030 and

0306-2619/$ - see front matter 2012 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.apenergy.2012.03.006

⇑ Corresponding author. Tel.: +45 45251699.E-mail address: [email protected] (Davide Tonini).

Applied Energy 99 (2012) 234–246

Contents lists available at SciVerse ScienceDirect

Applied Energy

j ou rna l homepage : www.e l sev ie r. com/ loca te / apenergy



2050. These future scenarios are compared with the existing en-ergy system in the reference year 2008. Environmental impactswere quantied using life-cycle assessment (LCA). All relevant en-ergy technologies and conversion processes in the energy systemwere addressed (e.g. wind energy, hydropower, photovoltaic, solarheating and Biomass-to-Energy (BtE) technologies); however, par-ticular focus was placed on the Biomass-to-Energy subsystems asbiomass and associated land use effects were a specic concern.

The specic objectives were: (i) identication of potential bio-mass resources, (ii) identication and selection of suitable biomassconversion technologies and related efciencies and (iii) quanti-cation of environmental impacts associated with the selected en-ergy scenarios.

2. Methodology

2.1. Goal, scope and functional unit

The goal of the LCA was to assess the environmental impacts re-lated to more sustainable energy scenarios in Denmark. A range of potentially future energy scenarios were selected based on a mix of residual agricultural resources, energy crops, wind and otherrenewables (e.g. waves and solar heating) and to the extent neededalso fossil fuels. All future energy scenarios were compared with areference representing the Danish energy system in 2008. The ser-vice dened by the LCA functional unit was ‘‘meeting the energydemand (electricity, heat and fuels) required in Denmark in theyears 2008, 2030 and 2050’’. This service does not describe theamount of energy provided, but it is dened to ensure that the fu-ture living standards are similar to the 2008 baseline. As the mod-eled energy demand and supply was not identical in the selectedscenarios, the LCA modeling results were normalized with the pri-mary energy supply (named ‘PES’) for the individual years to allowcomparison. This provided the intensity of the environmental im-pact per unit of primary energy supplied (e.g. Gg CO 2-eq/PJ PES) tothe system and allowed comparison of the energy systems in spiteof the different energy demands. For example, a decrease in envi-ronmental impact should be equal to or be larger than a decreasein supply in order to achieve an overall lower ‘normalized’ impact.The environmental impacts were quantied with a time horizon of 100 years according to common LCA practice [15] .

2.2. Assessment approach

The overall approach used for carrying out the life-cycle assess-ment included the following steps: (1) selection of potential energyscenarios based on available national energy strategies and politi-cal targets (e.g. regarding shares of wind power, CO 2 emissionreductions, etc.), (2) selection of relevant conversion technologies

and collection of associated technology data for the LCA, (3) bal-ancing energy supply and demand for each scenario to providethe necessary input for the LCA, (4) performance of the impactassessment and discussion of results.

Step (1) above was carried out based on energy system analysisof future 100% renewable Danish energy systems performed inseparate studies using the model EnergyPLAN [4–6,16] . Therefore,detailed documentation of the technical properties of the selectedenergy scenarios was outside the scope of this paper but can befound in the cited references.

2.3. Impact assessment

The life-cycle assessment was carried out according to theEDIP2003 methodology [17] for the environmental impact catego-ries: global warming, acidication and aquatic eutrophication

(distinguishing was made between nitrogen and phosphorous re-lated impacts). Impacts related to land occupation was includedin the assessment according to the IMPACT 2002+ methodologyfor this impact category [18] . For acidication, the results were ex-pressed as area of unprotected ecosystem within the full deposi-tion area that is brought to exceed the critical load of acidication as a consequence of the emissions (unit: 10 6 m 2/PJPES).Site-generic characterization factors (average values for EU15 plusSwitzerland and Norway) were used [17] . Site-dependent charac-terization factors were only available for a few compounds andthe values were very similar to the site-generic factors for Den-mark. Particularly, for SO 2, NO x and NH 3 the site-generic character-ization factors were 17.7, 8.6 and 23 m 2 of unprotected ecosystem/kg. For aquatic eutrophication, two sub-categories were usedaccording to the methodology: aquatic eutrophication (nitrogen)where impacts are expressed as kg N/PJ PES and aquatic eutrophica-tion (phosphorous) where impacts are expressed as kg P/PJ PES. Theemissions of N and P are accounted separately because there areaquatic ecosystems where the limiting nutrient is P (typically in-land waters in EU temperate regions, e.g. lakes) and others whereN is the limiting nutrient (e.g. marine waters). The relevant charac-terization factors for aquatic eutrophication were: 0.096 kg N/kgNO x, 0.1886 kg N/kg NH 3, 0.29 kg P/kg PO 3

4 (emitted to water),and 0.88 kg P/kg P (emitted to water).

For multiple-output processes such as bioreneries, wherevaluable byproducts (e.g. fodder or chemicals) were generated to-gether with fuels, system expansion was applied and it was as-sumed that these products substituted the marginal products inthe market, according to the principles of consequential LCA [19] .The life-cycle assessment was facilitated by the LCA software Sim-apro 7.1 [20] . The energy balances ( Figs. 1–3 ) were facilitated bythe software STAN [21] .

2.4. Energy scenarios

Five different energy scenarios were assessed: (I) ‘‘2008’’ (refer-ence), (II) ‘‘2030’’, (III) ‘‘2050CSV’’ (2050 conservative (CSV) sce-nario), (IV) ‘‘2050RME’’ (2050 scenario where biodiesel as rapemethyl ester (RME) is totally produced from rapeseed) and (V)‘‘2050BtL’’ (2050 scenario where biodiesel (FT-biodiesel) is pro-duced from lignocellulosic biomass through Biomass-to-Liquid(BtL) and Fischer–Tropsch (FT) technology, except for the shareof RME corresponding to the amount produced today in DK). Thelatter three represented different potential alternatives for trans-port fuel production in year 2050.

2.4.1. 2008 ScenarioThe ‘‘2008’’ scenario was selected as reference representing the

current energy system primarily based on fossil resources. Data for

energy demand and supply were based on Danish national statis-tics for 2008 [22] : the gross primary energy supply was 864 PJwhile the nal net consumption by society was 652 PJ (excludingtransmission losses). An overview of the energy scenario ‘‘2008’’is shown in Fig. 1.

Energy scenarios representing 2030 and 2050 were associatedwith signicant reductions in energy demand and based on im-proved efciencies of combined heat and power (CHP) plants, in-creased electricity production from wind energy, replacement of fossil fuels with biomass and the introduction of electric vehicles.Specications of the future energy scenarios were done based onavailable strategies and political targets published by the Danishauthorities [2,23] . A detailed overview of the technical measuresadopted is presented in Table 1 . Table 2 provides detailed datafor energy supply and demand for all energy scenarios while Ta-ble 3 provides data for the distribution of the energy consumption

D. Tonini, T. Astrup / Applied Energy 99 (2012) 234–246 235



among the different transportation means. Further informationregarding the energy modeling can be found elsewhere [4,5,16] .

2.4.2. 2030 ScenarioThe ‘‘2030’’ scenario represented a ‘‘link’’ between 2008 and the

2050 scenarios: more than 50% of the energy was generated from

renewable resources. The transition to electric passenger vehicleswas assumed to be incomplete and therefore ethanol was requiredas fuel in the energy system. The gross primary energy supply wasestimated to 679.4 PJ while the nal net consumption by societywas 551 (excluding transmission losses). The higher efciency of the energy system was mainly due to technical measures assumedto be implemented for reduction of the energy demand and supply[4] : e.g. decommissioning of old inefcient power plants, construc-tion of new more efcient power units (utilizing SOFC, i.e. SolidOxide Fuel Cells), implementation of geothermal units, reductionof electricity consumption in households (by 50% compared to2008) and in industry/services (by about 43% compared to 2008),improvement and expansion of district heating networks for cover-ing up to 70% of the heat demand, improved insulation of (old andnew) buildings and decrease of fuel consumption in industry(about 31% decrease compared to 2008). With respect to the trans-

port sector, the expected growth in demand (18%) was assumedcovered equally by train transport (assuming expansion of rail-roads and of high speed trains) and by avoiding transport throughintroduction of road pricing and improved urban planning. In 2030half of the passenger vehicles were assumed electric (or hybrid);the domestic ights were reduced to 5% of the current level by

substituting their capacity with high speed trains and the fuel de-mand for ships was decreased by 40% compared to 2008. Theimplementation of such measures in the energy analyses resultedin the energy scenario ‘‘2030’’ shown in Fig. 2. Industry requiredabout 115 PJ of ‘gases’ (as syngas and natural gas) and 13 PJ of elec-tricity to operate and produce heat required for the different indus-trial processes. Based on the gasication efciencies (seeSupporting Information , i.e. SI) to produce 85 PJ of syngas corre-sponding 117 PJ of biomass were needed. This was covered byavailable residual biomass resource ( Supporting Information ) andadditional 40 PJ of energy crops (willow). Willow was selected asa favorable energy crop among other options (e.g. Miscanthus , pop-lar, etc.) because of the high yield, low requirement of fertilizersand other agricultural practices, capacity of sequestering carbon,adaptability to different soils, etc. [24,25] . However, the choice of Miscanthus or other short rotation coppice (SRC) would not signif-

Fig. 1. Energy scenario ‘‘2008’’ (unit: PJ). The energy consumption of the offshore platforms for oil and gas extraction is included in the energy system (29 PJ). NG: natural gas,CO: coal, PV: photovoltaic, hydro: hydropower, E: electricity, H: heat, L: losses, BP: byproducts, F: fuel.

236 D. Tonini, T. Astrup / Applied Energy 99 (2012) 234–246



icantly affect the results, as yield and fertilizers needs for thesecrops are similar [25] . The electricity needed for transport, industry(and services), household and individual heat pumps, adding up toabout 123 PJ (including transmission losses) was produced from

wind (84 PJ), hydropower/photovoltaic (8.3 PJ), CHP plants usingsyngas and a large share of methane (from anaerobic digestion of manure and grass) in SOFC (about 22.5 PJ), from combustion of oil, coal, waste and byproducts from biomass conversion (about4 PJ) and industry (4.6 PJ). For heating, 180 PJ of heat were re-quired; these were produced from solar heating (19 PJ), heatpumps (57 PJ), CHP plants using syngas, methane in SOFC (15 PJ),district heating plants burning coal, waste, byproducts from bio-mass conversion (79 PJ) and industry (9.5 PJ). In Fig. 2, the electric-ity needed for industrial heat pumps was taken into account asdecreased electricity delivered to the net from the ‘power units’.For transportation (details in Table 3 ), 0.8 PJ of methane, 72 PJ of diesel, 4.5 PJ of RME-biodiesel, 27.5 PJ of petrol, 4.7 PJ of bioethanoland 37.9 PJ of aviation fuel were required. Lastly, 33 PJ of fossilfuels were needed to operate offshore platforms for extraction of oil and natural gas.

2.4.3. 2050 ScenariosFor 2050, three different versions of the energy scenario were

assessed. The three versions represented three fundamentally dif-ferent approaches for production of transport fuels (all other as-

pects of these three alternatives (industry, power plants,household) were identical, see Fig. 3). In order to fulll the energydemand, about 51 PJ of willow were required to be cultivated inaddition to the estimated potential biomass resources. Most of the transportation was based on electricity produced from renew-ables (also identical in the three scenarios), except for heavy vehi-cles and aviation which required diesel and long-chainhydrocarbons (kerosene and aviation fuel). The nal net energyconsumption by society was 535 PJ (excluding transmissionlosses). Technical measures, similar to those in 2030, for reductionof the overall energy demand and supply were included ( Table 1 ).These included: increase of renewables from wind, hydro and so-lar; further reduction of fuel consumption (33% compared to2008) in industry; ‘avoided’ passenger growth compared to 2030(by implementing the same measures as in 2030); electricationof terrestrial transportation (now relying 100% on electric vehicles)

Fig. 2. Energy scenario ‘‘2030’’ (unit: PJ). The energy consumption of the offshore platforms for oil and gas extraction is included in the energy system (33 PJ). NG: natural gas,GH: geothermal heating, ST: solar heating, PV: photovoltaic, Hydro: hydropower, CO: coal, Syn: syngas, BG: biogas, BP: byproducts, E: electricity, H: heat, H : process heat(industry), L: losses, PE: petrol, DS: diesel (including 4.5 PJ of RME-biodiesel), BE: bioethanol, AF: aviation fuel.




Fig. 3. Energy scenarios ‘‘2050CSV’’, ‘‘2050RME’’ and ‘‘2050BtL’’ (unit: PJ). The energy consumption of the offshore platforms for oil and gas extraction is included in theenergy system (10.5 PJ in ‘‘2050CSV’’; 4.8 PJ in ‘‘2050RME’’ and ‘‘2050BtL’’). GH: geothermal heating, ST: solar heating, PV: photovoltaic, Hydro: hydropower, NG: natural gas,Syn: syngas, BG: biogas, BP: byproducts, E: electricity, H: heat, H : process heat (industry), L: losses, DS: diesel, BD: biodiesel, AF: aviation fuel, SF: synthetic fuel.




and improved efciency of the ships (fuel demand lowered by 60%compared to 2008). These measures resulted in the energy owsshown in Fig. 3. The electricity needed for transport, industry(and services), households and individual heat pumps, adding upto about 180 PJ (including transmission losses) was produced fromwind (133 PJ), hydropower/photovoltaic (25 PJ), CHP plants usingsyngas and a large share of methane (from anaerobic digestion of manure and grass) in SOFC (about 10 PJ), combustion of wasteand byproducts from biomass conversion (about 10 PJ) and indus-try (3.6 PJ). About 178.6 PJ of heat were required for householdsand industry/services heating. These were produced from solarheating (19 PJ), heat pumps (57 PJ), CHP plants using syngas, meth-ane in SOFC, district heating plants burning waste and byproductsfrom biomass conversion (overall 93 PJ) and industry (9.3 PJ). Com-pared to 2030, the fuel (as gas) needed to industry was reduced to85 PJ. The remaining ‘process heat’ demand was covered with heat

pumps (determining an electricity input from renewable intermit-tent sources, e.g. wind, of about 34 PJ). Instead, the production of the biodiesel for heavy terrestrial transport (e.g. lorries), shipsand defence varied depending on the scenario. In ‘‘2050CSV’’ (i.e.2050 conservative scenario), 63 PJ of crude oil was assumed to ful-ll the demand for biodiesel (about 35 PJ) and aviation fuel (about33 PJ). The gross primary energy supply was 559 PJ. In ‘‘2050RME’’(i.e. 2050 scenario where biodiesel (as RME) is entirely producedfrom rapeseed), 35 PJ rape methyl ester (RME) was assumed to ful-ll the fuel demand of biodiesel supplemented by 33 PJ of crude oilfor aviation. The gross primary energy demand was 576 PJ. In‘‘2050BtL’’ (2050 scenario where biodiesel is produced from ligno-cellulosic biomass through BtL- and FT-technology, except for theshare of RME corresponding to the amount already produced todayin DK), 30.5 PJ of FT-biodiesel from willow and 4.5 PJ of RME wasassumed for terrestrial transportation, ships and defence

Table 1

Overview of the technical measures adopted in the energy scenarios for 2030 and 2050 to decrease energy demand. A detailed energy system analysis of the scenarios can befound in [4,5,16] . The resulting energy balance with respect to heat, electricity and transport fuel production is shown in Figs. 2 and 3 and Tables 2 and 3 . SOFC: Solide Oxide FuelCell; CHP: combined heat and power.

Energy sector Scenario 2030 Scenarios 2050 Effect Notes

Electricity production Introduction of SOFC CHPs (33% of total CHP plants).

Introduction of SOFC CHPs (100%of total CHP plants).

Reduction of primary energysupply from fossil fuel.

Decommission of old power plants.

Geothermal plants are installedreaching a total production of about 4 TW h (15% of demand inbig cities).

Integration of geothermal withincineration plants.

Electricityconsumption

Electricity demand in household isdecreased by 50% compared to2008.

Reduction of electricityconsumption and of primaryenergy supply from fossil fuel.

Installation of best-practiceproducts, increase of productsstandards and use of informationcampaign.

The electricity demand in industryand service is reduced by about43% compared to 2008.

District cooling is implementedaccounting for half of the potentialsavings in 2030 [44] .

Heat production The level of net coverage fromdistrict heating is increased to 70%(46% in 2008).


Expansion of district heatingnetwork [45–47] .

Heat consumption New buildings use 75% of thedemand used in 2008 by newbuildings.

Reduction of heatconsumption and of primaryenergy supply from fossil fuel.

Improvement of insulation [45–47] .

Old buildingsachieve a heat savingof 50% of the current heat demand.

Improvement of insulation [45–47] .

Industry (process heat) The fuel consumption in industryand services is reduced by 31% of 2008 value.

The fuel consumption inindustry and services is reducedby 33% of 2008 value.


Combination of end use savings,more efcient technologies andinstallation of industrial heatpumps [48] .

Terrestrial transport The expected growth in passengertransport (18%) is assumedcovered equally by train transportand by ‘avoiding transport’(introduction of road pricing andimproved urban planning).

Growth in passenger transport(compared to 2030 value) is‘avoided’ (introduction of roadpricing and improved urbanplanning).

Reduction of fuelconsumption and of primaryenergy supply from fossil fuel.

Introduction of road pricing andbetter physical urban planning.

Vehicles meet 60% of the transportdemand. Railroads meet 30% of thetransport demand.

Vehicles meet 50% of thetransport demand. Railroadsmeet 40% of the transportdemand.

Expansion of railroads (highercapacity and higher speed) andimplementation of better physicalurban planning [49,50] .

Bicycling and walking meet 10% of the transport demand (5% in2008).

Bicycling and walking meet 10%of the transport demand (5% in2008).

Installation of loading stations;electric vehicles are also needed asstorage system to accommodateelectricity uctuations.

Bioethanol covers about 5% of thepassenger transport.

No bioethanol is needed fortransportation.

Half of the good transport isdiverted to ships and trains.


Introduction of road pricing andexpansion of railroads.

Aviat ion Domest ic ights a re reduced to 5%of the current demand.


Expansion of railroads andimprovement of fuel-combustiontechnology of the planes.

Shipping Ships will lower the fue l demandby 40% compared to 2008.

Ships will lower the fuel demandby 60% compared to 2008.


Improvement of fuel-combustiontechnology of the ships.




supplemented by 33 PJ of crude oil for aviation. The gross primaryenergy demand was 588 PJ. It has to be noted that in Figs. 2 and 3 ,industry and electrolysis units are part of the energy system (theyutilize electricity to produce other energy carriers). This should berealized when comparing data in Tables 1 and 2 with Figs. 2 and 3 .

2.5. Life-cycle inventory data

2.5.1. Biomass resourcesThe relevant biomass resources available in Denmark were:

manure, grass, lignocellulosic biomass (e.g. wood and straw) andwaste. The total amount of biomass potential was estimated to

be about 182.3 PJ. Focus was on residual biomass, i.e. waste (e.g.municipal solid waste, MSW) and byproduct/residues from agricul-ture and forestry (e.g. straw, manure, wood). Today most biomassresources have a function in the ecosystem or in the economymeaning that the utilization of these resources for energy produc-tion would induce changes in the ecosystem or in the society if sta-tus quo is to be maintained. As a consequence, the use of biomassresources for energy purposes instead of the current use (e.g. feed-ing, bedding, ploughing back to elds, etc.) will nally lead to acompetition between energy and other uses. The consequences of routing biomass resources to energy production were addressedin the LCA. Municipal solid waste quantities currently incinerated

Table 2

Energy supply and demand in the individual energy scenarios years (PJ; rounded values): 2050CSV (conservative, fossil diesel used for heavy terrestrial transport), 2050RME(RME-biodiesel used for heavy terrestrial transport), 2050BtL (FT-biodiesel mainly produced through BtL for heavy terrestrial transport). Primary energy represents the amount of energy supplied to the energy system with biomass, fossil fuels and renewables. ‘Tot’ refers to the nal energy (sum of electricity, heat, fuel and ‘non-energy’ use) delivered toconsumers in society (three values are provided with and without including industry, electrolysis and transmission losses).

2008 2030 2050CSV 2050RME 2050BtL

Energy supplyFossil fuel a 704 283 87 40 40Primary energy (total) 864 679 559 576 588

Energy demand (electricity)Household 37 18.4 18.4 18.4 18.4Individual heat pumps 1.8 6.1 5.8 5.8 5.8Industry/services 83 48 46 46 46Industrial heat pumps – 13.1 34 34 34Other b – – 3.2 3.2 3.2Transport sector 1.4 24.1 43.3 43.3 43.3Electrolysis units – – 27 27 27Tot (exc. industry, electrolysis, transm. loss) c 40 97 111 111 111Tot (inc. industry, electrolysis; exc. transm. loss) 123 110 171 171 171Tot (incl. industry, electrolysis and transm. loss) 131 119 180 180 180

Energy demand (heat)Household 147.4 147 147 147Industry (process heat) 114 106 106 106Industry/services 33.5 31.6 31.6 31.6Total 298 d 294 284 284 284

Fuel for transport 220 147 80 80 80

’Non-energy’ use 11 – – – –

Tot (exc. industry, electrolysis, transm. loss) 569 538 475 475 475Tot (inc. industry, electrolysis; exc. transm. loss) 652 551 535 535 535Tot (incl. industry, electrolysis and transm. loss) 660 560 544 544 544

a The share of fossil fuel in the waste (e.g. plastic) is excluded from the values provided.b Sum of ethanol, biogas production, district cooling and electrical cartridges. In 2050 this was modeled as a negative consumption as it provided net electricity to the

energy system [4,5,16] .c Net electricity delivered to nal consumers. When comparing with the values in Figs. 1–3 , it has to be kept in mind that in the gures the electricity (from renewables)

consumed by industry and electrolysis units is visualized as internal ow as industry and electrolysis units are part of the energy system (i.e. they use electricity to generateother energy carriers). In Figs. 1–3 the transmission losses are incorporated in the ow ‘‘loss’’ to simplify the sankey diagram. This also applies to the heat ows.

d Calculated as difference between the total primary supply and the sum of: electricity, fuel and ‘non-energy’ use. The heat from district heating corresponded to 124 PJ; theremaining (174 PJ) is produced by combustion of fossil fuels and biomasses in boilers.

Table 3

Distribution of energy consumption for transportation in 2030 and 2050 (PJ): E (electricity), BD (biodiesel), DS (diesel), PE (petrol), bioethanol (BE), synthetic fuel (SF), methane(CH4) and aviation fuel (AF).

Transport 2030 2050

E BD DS PE BE CH4/SF AF E BD CH4/SF AF

Passenger cars 10.7 1.3 5.7 27.5 4.7 – – 17.1 – 5.1 –Vans 5.9 3.2 18.5 – – – – 7.4 – 2.2 –Buses – – 9.1 – – 0.8 – 0.9 5.7 1.1 –Lorries – – 33.6 – – – – – 25.2 3.4 –Passenger trains 6.6 – 0.2 – – – – 16.9 – – –Freight trains 0.9 – – – – – 1 – – –Domestic aviation – – – – – – 0.1 – – – 0.1International aviation – – – – – – 37.8 – – – 33.4Ship transport – – 2.8 – – – – – 1.8 – –Defence – – 2.1 – – – – – 2.1 – –Total 24.1 4.5 72 27.5 4.7 0.8 37.9 43.3 34.8 11.8 33.5

(171.5) (123.4)




(40 PJ) were estimated to increase to 47 PJ in 2030 and 2050 (noincreases were assumed between 2030 and 2050 because addi-tional recycling was anticipated). A detailed description of the bio-mass potential for Denmark is reported in the SupportingInformation (SI).

2.5.2. Energy conversion technologiesSelection of the BtE conversion technologies was based on a

number of considerations which implied energy system as wellas technical issues for handling the biomass (see SI). The produc-tion of an intermediate energy carrier (e.g. biogas and syngas)was preferred to direct combustion for the exibity and storabilityof the energy products which are needed to accommodate the uc-tuations of energy systems with high penetration of wind power[26] . Manure and grass were assumed to be fermented to biogasthrough anaerobic digestion processes. Lignocellulosic biomass(e.g. wood, straw and willow) was assumed to be gasied for syn-gas generation. Biogas and syngas were then converted to heat and

electricity in Solid Oxide Fuel Cells (SOFCs) with high electricityefciency (electricity efciency 54% and heat efciency 36%). In2008 and 2050, MSW was assumed to be incinerated for heatand electricity production. In the 2030 scenario a share of theMSW (plastic) was however gasied to cover gas demands inindustry. Biodiesel was produced from rapeseed and willow bymeans of transesterication and thermal process (gasicationand Fischer–Tropsch), respectively.

Table 4 provides an overview of the background life-cycleinventory (LCI) data for (selected) Biomass-to-Energy (and to-Fuel)processes used in the assessment. With respect to the LCIs for:wind, hydro and wave power, heat pumps, SOFC, fossil fuel com-bustion in combined heat and power (CHP) plants, district heatingplants, peak-load boilers, vehicles, offshore platforms and indus-trial furnaces for heat production, common processes found inthe Ecoinvent database [27] were used. A detailed description of

the energy conversion technologies is reported in the SI.

2.5.3. Land use changesCultivation of energy crops requires use of land thereby induc-

ing direct and indirect land use changes (dLUCs and iLUCs) underthe basic assumption that land available for cultivation isconstrained.

With respect to willow, Soil Organic Carbon (SOC) changes anddLUC were estimated based on [28] . The iLUC were estimatedbased on the assumption that expansion of willow cultivated landin Denmark replaced the marginal crop (spring barley) which hadto be produced somewhere else if status quo was to be maintained.The most likely consequence was assumed to be conversion of grassland into barley (69%) as well as intensication of barley cul-tivation in Canada (31%) [29,30] . The land use consequence (interms of change in SOC stock) of replacing prairie grass with barley

was 84 Mg CO 2/ha. Intensication implied a larger utilization of fertilizers in order to increase the production on the same con-strained land [29,30].

With respect to rapeseed, SOC changes, dLUC and iLUC werequantied according to [30] assuming conversion of set-aside landinto rapeseed (all 2050 scenarios) or conversion of set-aside landand arable land (spring barley) into rapeseed (only the‘‘2050RME’’ scenario). For conversion of arable land (spring barley)into rapeseed, a SOC loss of 0.115 Mg C/ha/y was assumed (see SI).Only dLUC and iLUC associated with changes in rapeseed cultiva-tion from the current situation to the future needs wereconsidered.

It has to be noted that at the time the current study was started(2008) the research on the impacts related to dLUC and iLUC wasstill at an early stage and the literature and knowledge availablelimited and subject to continuous developments. Therefore, themethodology as well as the nal estimations of dLUC and espe-cially iLUC was characterizedby signicant uncertainty. The uncer-tainties in the assumptions were addressed (as best as possible) inthe sensitivity analysis. Table 5 provides an overview of the back-ground data used to evaluate the SOC changes (and related dLUC)associated with changes in the use of land. A detailed discussion of the impacts associated with land use changes is reported in theSupporting Information .

2.5.4. Management of agricultural and biomass conversion residualsThe removal of straw from elds induces changes in the soil car-

bon stock. The calculated carbon depletion was 0.09 Mg C/Mgstraw [31] . Removal of nutrients (N, P and K) with the straw ledto additional fertilizer use to maintain constant crops yields. Strawremoval also induced to decrease N 2O emissions: a decrease of 0.03 kg N–N 2O/Mg DM straw was assumed based on [31] .

The use of grass for energy instead of feeding induced an in-creased demand for other types of fodder. This was modeled with

additional production of barley in order to satisfy the feed demand.

Table 4

Overview of background data used for the LCA in relation to energy conversion technologies for selected Biomass-to-Energy (or -Fuel) processes: Biomass-to-Liquid process (BtL),Fischer–Tropsch (FT), BE (bioethanol).

Biomass Energy technology Products Use of products and byproducts Reference

Manure Anaerobic digest ion Biogas and digestate Biogas to hea t and e lect rici ty. Diges ta te to use on land [27,32,33]Grass Anaerobic digestion Biogas, solid biofuel and

proteinsBiogas to heat and electricity. Grass bers to heat andelectricity. Proteins substitute soymeal

[27]

Wood and willow Gasica tion Syngas and b iochar Syngas to hea t and e lectr ic ity. Biochar to use on land [27,51–53]Straw Gasication Syngas and biochar Syngas to heat and electricity. Biochar to use on land [27,54]Waste Incineration Electricity and heat – [27,55]Rapeseed Transesterication RME, glycerin and solid

biofuelRME to transport. Glycerin substitutes glycerin production.Biofuel to heat and electricity

[27]

Willow (BtL, FT) Gasication and Fischer–Tropsch FT-biodiesel and biochar FT-biodiesel for transport. Biochar to use on land [56]Straw (BE) Straw renery BE, molasses and solid

biofuelBioethanol for transport. Molasses substitutes fodder. Biofuelto heat and electricity

[27,57]

Table 5

Overview of background data for SOC changes used in the LCA in relation to effectsassociated with direct land use changes (dLUCs) for selected crops. Positive valuesindicate emissions (e.g. loss of carbon) while negative values indicate sequestration/avoided emissions (1 ha = 10 6 m 2). The values express the net ‘‘delta’’ betweenemissions in the new crop-system and emissions in the replaced system. ⁄ Temporalscope for the operation of the system: 20 years.

Crop system Mg CO 2 /ha Mg N 2O /ha Mg NO 3 /ha Reference

Grassland to springbarley

84 0.02 4.2 [30]

Set-aside land torapeseed

88 0.022 4.6 [30]

Spring barley to willow 55 ⁄ 0.0033 ⁄ 4 ⁄ [28]




The use on land of digestate from anaerobic digestion of manurewas credited by substitution of inorganic N, P, K fertilizers [32,33] .Application of 1 Mg of digestate was assumed to substitute 4.07 kgof ammonium nitrate (as N), 2.1 kg of triple superphosphate (asP2O5) and 3.3 kg of potassium chloride (as K 2O). The higher amountof N substituted (compared to direct application on land of rawmanure) was a consequence of the higher availability of N in thedigestate after the selected manure treatment. This has been thor-oughly discussed in [32,33] .

The use on land of biochar was also credited for its potential po-sitive effects on soil, e.g. carbon sequestration, improved fertilizerefciency and reduced N 2O emissions, based on [34] . A detaileddescription of the assumptions regarding management of agricul-tural and biomass conversion residuals is reported in the Support-ing Information .

3. Results and discussion

The results are presented with respect to the environmental im-pact categories: global warming (GW), acidication (AC), aquaticeutrophication, nitrogen (EP (N)) and land occupation (LO) inFig. 4. Only the sub-processes (e.g. transportation, LUC, fossil fuelcombustion, etc.) contributing to the overall impacts with morethan 1% are shown. For eutrophication, distinction was made be-tween impacts caused by nitrogen emissions (EP (N)) and eutro-phication caused by phosphorous emissions (EP (P)). For thepurpose of clarity, only results for EP (N) are included in Fig. 4.

The results for aquatic eutrophication related to phosphorous (EP(P)) were similar in the trend to those of nitrogen, EP (N), andare presented only in Table 6 .

The results for the category land occupation are presented asadditional land required (10 6 D m 2/PJPES) compared with the cur-rent situation (‘‘2008’’). For acidication, the results are expressedas area of unprotected ecosystem that exceeds the critical load of acidication as a consequence of the emissions (i.e. 10 6 m 2/PJPES).The impacts were calculated per unit of primary energy supplyprovided to the energy system (i.e. PJ PES) (e.g. Gg CO2-eq/PJ PES).This provided the intensity of the environmental impact and al-lowed comparison of scenarios with different primary energysupply.

To enable a more direct comparison between the individual en-ergy scenarios, Table 6 provides both normalized and total envi-ronmental impacts for the energy scenarios. Additionally, ‘‘total’’values (‘‘Total REF’’) were calculated for each energy scenario un-der the assumption that the primary energy supply was identicalto the reference year 2008 (864 PJ). The intention with these valueswas to illustrate the effects from decreasing energy demand vs. theeffects from changing the energy supply.

Fig. 5 shows the environmental impacts associated with theproduction and combustion of the transport fuels RME and FT-bio-diesel compared with traditional diesel. In this case the results pre-sented in Fig. 4 were re-calculated in order to correspond to afunctional unit of 1 energy unit of diesel-fuel. The intention withthis calculation was to clarify the environmental impacts relatedto the individual fuels for better comparison.

Fig. 4. Contribution of the different sub-processes to the impact on the selected environmental categories. The impacts are normalized with the primary energy supplied(‘‘PES’’) to the individual energy scenarios. PP: power plants; LUC: sum of dLUC and iLUC; BtE: Biomass-to-Energy.




3.1. Global warming (GW)

Overall, the results for GW indicated decreasing intensity of theGHGs emissions per unit of primary energy supply from 2008(about 68 Gg CO 2-eq/PJ PES) to 2050 (about 17–31 Gg CO 2-eq/ PJ PESdepending on the scenario). The reduction of GHGs emissions(per PJ PES) was thus in the range of 66–80%. This was primarilyattributed to the substitution of fossil fuel with biomass resources,the increased wind energy penetration in the system and to theconversion of transport to electric passenger vehicles. Only partialconversion of transport to electric vehicles in ‘‘2030’’ (still relyingon fossil fuel for 50% of passenger vehicles) explained the higher

impacts compared to 2050s where electrication of passengervehicles was completed. The total GHGs emissions decreased from59 Tg in ‘‘2008’’ to 26 Tg in ‘‘2030’’ and 10–18 Tg in the 2050s(depending on the scenario). It has to be noted that the total GHGsemissions calculated for 2008 (59 Tg) is higher than the value(48.4 Tg) indicated in the Countries ofcial statistics by [35] as thislatter one only counted for emissions associated with the fuelscombustion (i.e. upstream emissions associated with fuel produc-tion and provision were not included).

The decrease on the total GHGs emissions was both attributedto the decreased impact intensity (impact per PJ PES) and to thediminished energy demand. As shown in Table 6 , if the total energy

Table 6

Overview of energy supply in the individual energy scenarios (fossil fuel and total primary energy supply) and of environmental impacts: (1) normalized per primary energysupply, (2) total impacts for each scenario, and (3) potential totals assuming that primary energy supply was identical to the reference year 2008 (864 PJ).

Unit Energy system

2008 2030 2050CSV 2050RME 2050BtL

Energy supplyFossil fuel PJ 704 283 87 40 40Primary energy (total) PJ PES 864 679 559 576 588

ImpactsGW Normalized Gg CO 2-eq/PJ PES 68 38 20 31 17

Total Tg CO2 59 26 11 18 10Total REF Tg CO2 59 33 17 27 15

AC Normalized 10 6 m 2/PJPES 301 193 121 177 135Total 10 9m 2 260 130 68 100 79Total REF 109 m 2 260 167 105 153 117

EP (N) Normalized Mg N/PJ PES 14 16 15 29 24Total Gg N 12 11 8.4 17 14Total REF Gg N 12 14 13 25 21

EP (P) Normalized Mg P/PJ PES 0.1 2.0 2.4 3.9 2.8

Total Gg P 0.1 1.3 1.3 2.3 1.7Total REF Gg P 0.1 1.7 2.1 3.4 2.4

LO Normalized 10 6 D m 2/PJPES – 624 861 2089 1787Total 10 9 D m 2 – 420 480 1200 1100Total REF 109 D m 2 – 539 744 1805 1544

Fig. 5. Environmental impacts associated with production and combustion of biodiesel (and fossil diesel) for heavy terrestrial transportation, ships and defense (functionalunit: 1 PJ diesel-fuel). Excluding potential environmental benets associated with the use on land of biochar from gasication processes.




supply was the same as ‘‘2008’’ (864 PJ), the total impact for GWwould still be signicantly lower than today. Although the energymix (i.e. share of electricity, heat and fuel) differed in the selectedperiods compared to ‘‘2008’’, Table 6 demonstrates that the bene-ts were primarily attributed to the substitution of fossil fuels withbiomasses, the increased wind penetration and the conversion toelectric passenger vehicles, rather than to the diminished energydemand (and consequent supply).

The preferred scenarios from a GW perspective were the‘‘2050BtL’’ and ‘‘2050CSV’’ scenarios whereas the worst was the‘‘2050RME’’ scenario. The difference among the 2050 scenarioswas caused by the magnitude of iLUC impacts associated with en-ergy crops cultivation for biodiesel production to cover the de-mand for heavy terrestrial transportation, ships and defense. Theimpacts associated with LUC (i.e. sum of dLUC and iLUC) were esti-mated to about 5, 20 and 8 Gg CO 2-eq/PJ PES in the ‘‘2050CSV’’,‘‘2050RME’’ and ‘‘2050BtL’’ scenarios, respectively. The impactsassociated with rapeseed cultivation in ‘‘2050RME’’ were signi-cantly higher than those for willow in ‘‘2050BtL’’ due to the loweryield of rapeseed and hence higher iLUC. In this context, the use of (traditional fossil) diesel for heavy transport, ships and defensewas still favorable over RME, whereas FT-biodiesel production(the ‘‘2050BtL’’ scenario) showed slightly lower GW impacts thanfossil diesel. Although the results for FT-biodiesel strongly de-pended on the assumptions regarding biochar effects and willowyield, this demonstrated that the considerable iLUC’s associatedwith cultivation of energy crops can completely off-set the benetsof biofuels (see further discussion of biofuels in Section 3.3 ). Lastly,it should be noted that use on land of digestate and biochar led tosignicant GW savings due to the return of nutrients and carbon tothe soil.

3.2. Acidication (AC), aquatic eutrophication (EP) and landoccupation (LO)

The results for AC followed the trends observed for GW. De-creased NO x and SO x emissions from fossil fuel combustion inpower plants lowered the intensity of the acidication impactcompared to ‘‘2008’’. The best scenario was ‘‘2050CSV’’ contribut-ing with a load of 121 10 6 m 2/PJPES, while the ‘‘2050BtL’’ scenarioat 135 10 6 m 2/PJPES was second best. In the ‘‘2050BtL’’scenario,the environmental load was mainly associated with tailpipe emis-sions of NO x from biodiesel combustion in heavy vehicles and ships(corresponding to about 58 10 6 m 2/PJPES). Biodiesel-fueled heavyvehicles generally have higher NO x emissions than conventionaldiesel-fueled vehicles [36–39] . A similar situation was also the casefor ‘‘2050RME’’ where biodiesel for heavy transport and ships wasproduced from rapeseed. Among the 2050 scenarios, the worstenvironmental performance for AC was observed for ‘‘2050RME’’(177 10 6 m 2/PJPES) where cultivation of rapeseed contributed

with 34 106

m2/PJPES (N-fertilizers) in addition to the tailpipe

NO x related impacts. The scenario ‘‘2030’’ had higher impacts than2050s due to higher consumption of fossil fuels in the power andtransport sectors. In the 2050 scenarios, fossil fuels were largely re-placed with wind power. For all the selected scenarios, the totalacidication impacts were signicantly lower than today’s due toboth the decreased impact intensity and to the reduced primaryenergy supply (the latter one as a consequence of the reduced de-mand) ( Table 6 ).

For aquatic eutrophication (N and P), all the assessed scenarioscontributed with signicant impacts, mainly associated with theincreased use of fertilizers for energy crop production and the in-creased use on land of digestate from anaerobic digestion of grassand manure, with consequent potential release of nitrates andphosphates to surface waters. For both eutrophication categories,the least preferable scenario was ‘‘2050RME’’ (29 Mg N/PJ PES and

3.9 Mg P/PJPES): the potential eutrophication impact related tonitrogen was doubled whereas the impact associated with phos-phorous increased by one order of magnitude compared with‘‘2008’’. This was due to the large amounts of fertilizers requiredfor rapeseed and barley cultivation as a consequence of the cascad-ing effects associated with replacement of the marginal crop inDenmark (spring barley). A signicant increased in phosphorousemissions was observed as a consequence of (i) application on landof digestate (all 2030 and 2050 scenarios) and (ii) energy crops cul-tivation with related cascading effects (primarily for ‘‘2050RME’’due to rapeseed cultivation); this increase represents a potentialproblem for inland water ecosystems where P is typically the lim-iting nutrient for algae and plants growth.

Cultivation of willow in ‘‘2050BtL’’ required signicantly lessfertilizers than cultivation of rapeseed in the ‘‘2050RME’’ scenario,thereby causing lower impacts related to aquatic eutrophication.This was in agreement with several other studies, e.g. [40] . TheEP impacts associated specically with transportation was highestin the 2050 scenarios including biodiesel-fueled heavy vehicles(‘‘2050RME’’ and ‘‘2050BtL’’) because of higher NO x tailpipe emis-sions, also in accordance with the results for AC.

The ‘‘2050RME’’ and ‘‘2050BtL’’ scenarios required the largestarea of land (additional2089 and 1787 10 6 m 2/PJPES, respectively,compared with ‘‘2008’’). This was caused by cascading effects dueto the cultivation of energy crops in Denmark and subsequent dis-placement–replacement mechanisms as previously mentioned.The scenarios ‘‘2030’’ and ‘‘2050CSV’’ required signicantly lessadditional land due to use of fossil fuels for heavy terrestrial trans-portation, ships and defense in place of biodiesel.

3.3. Impacts for biodiesel production

Among the three evaluated scenarios for diesel-like fuel produc-tion (needed for heavy terrestrial transport, ships and defense),

RME was by far the least desirable option with respect to all envi-ronmental impact categories. With respect to GW, the impact wasestimated to 287 Gg CO 2-eq/PJ of fuel, whereas for fossil diesel thecorresponding value was about 89 Gg CO 2-eq/PJ of fuel. These re-sults clearly illustrate the importance of the large upstream GWimpacts related to land use changes (LUCs) and are in agreementwith other ndings in literature (e.g. [41,42] ).

The impacts from FT-biodiesel were in the range of 65–88 GgCO2-eq/PJ of fuel depending on assumptions regarding benetsfrom biochar (see sensitivity analysis). RME-biodiesel was alsothe least favorable option in relation to the AC, EP and LO catego-ries. With respect to AC and EP (N and P), the loads were mainlyassociated with tailpipe emission of NO x and use of N and P fertil-izers for crop cultivation, as previously explained. This was also thecase for FT-biodiesel. However, the impacts for AC and EP (N and P)associated with cultivation of willow for FT-biodiesel production

was signicantly lower than RME due to the higher yield and re-duced fertilizer use.

Fig. 5 shows the potential environmental impacts associatedwith diesel and biodiesel production. The impacts for the categoryEP (P) (not shown in Fig. 5) equaled 0.1, 28 and 8 Mg P/PJ respec-tively for diesel, RME- and FT-biodiesel.

3.4. Sensitivity analysis

The sensitivity of the results towards changes in assumptionsand parameters was carried out in order to assess the signicanceof: (1) willow yield, (2) magnitude of iLUC associated with cultiva-tion of rapeseed, (3) efciency of the BtL processes and (4) biochareffects on GW. Acknowledging the level of uncertainty associatedwith parameter selection, these aspects were identied by randomparameter variation as having the largest potential for affecting the




overall conclusions. (1) The yield of willow (baseline value:11.8 Mg DM/ha) was varied between 7 and 16 Mg DM/ha whichis a likely range for Denmark [43] . (2) The iLUC associated with ra-peseed (baseline values in Table 5 ) was estimated according to [41]thereby decreasing the impacts compared with the baseline sce-narios. An average value of 1.32 kg CO 2/kg rapeseed was assumed(only effects related to GW was assessed as no data for impacts inother impact categories were available). (3) The efciency of theBtL process (baseline value: 40%) was set to 57% with use of hydro-gen generated from wind power electrolysis of water (excess windpower was assumed to be available). (4) No benets for GW frombiochar were assumed (for the baseline assumptions see SI). Re-sults from the sensitivity analysis are presented in Table 7 .

Assuming a lower yield for willow (1) made the use of fossil fuelfor heavy transport favorable to FT-biodiesel produced from ligno-cellulosic biomass, according to all impact categories. Conversely,increased yield would be benecial for all impact categories, espe-cially with respect to the ‘‘2050BtL’’ scenario where willow is usedas substrate for FT-biodiesel production. A new value for iLUC (2)changed the overall result for GW for the ‘‘2050RME’’ scenario.The impacts associated with iLUC signicantly decreased(D = 11 Gg CO2-eq/PJ PES) compared with the baseline scenario.This difference however did not affect the overall ranking of the2050 scenarios: the scenario based on RME was still the leastfavorable.

More efcient thermochemical processes (3) combined withelectricity supply from wind power only slightly improved theenvironmental performance of biofuel production via thermo-chemical conversion. Although several technologies for utilization

of excess wind power in future energy systems will exist, it shouldbe noted that constraints (e.g. capacity and interconnectors) in theelectricity system may be limiting utilization of this electricity andthat local storage/utilization technologies may be needed. Regard-ing biochar (4), the performance of the ‘‘2050BtL’’ scenario becamesimilar to the ‘‘2050CSV’’ scenario if carbon sequestration from bio-char was not included, i.e. FT-biodiesel did not contribute with sav-ings in the GW category compared with fossil diesel.

Overall, the sensitivity analysis revealed that the quanticationapproach for LUC impacts, the assumption regarding yields of en-ergy crops and potential benets from biochar use can signicantlyaffect the overall result of the LCA. However, despite these effects,the overall ranking of the individual energy scenarios did notchange. The overall results based on the assessed scenarios aretherefore considered robust and not sensitive towards changes inassumptions.

4. Conclusion

The environmental impacts related to four potentially futureenergy scenarios for Denmark were compared with the energy sys-tem in 2008 by means of LCA. It was demonstrated that: (1) signif-icant reductions in GHGs emissions and global warming impactscan be achieved per PJ of energy supplied, (2) residual domesticbiomass resources were insufcient to cover demand for biomassenergy thereby requiring additional cultivation of energy cropswhich caused signicant environmental loads in most impact cat-egories, (3) large impacts associated with upstream land usechanges (LUCs) made the use of fossil diesel for heavy transportfavorable to RME- and FT-biodiesel, and (4) high potential aquaticeutrophication effects were a direct consequence of energy cropscultivation. Reduction of the energy demand, increased share of wind power and replacement of fossil fuels with residual domesti-cally available biomasses represented the main means for GHGsemissions savings in the future energy scenarios. However, by farthe main ‘‘environmental challenge’’ was the supply of biofuelsfor heavy terrestrial transport, ships, defense and aviation. Use of energy crops to fulll this demand caused signicant environmen-tal impacts related to global warming (mainly due to LUC), aquaticeutrophication (increased fertilizers use) and land occupation.Consequently, use of fossil diesel for these applications appearedpreferable over the biomass based fuels, except for global warmingwhere FT-biodiesel performed slightly better. The recommenda-tion, therefore, is to focus on residual domestically available bio-masses and minimize energy crops production.

Acknowledgments

The authors acknowledge the inputs and contributions fromBrian Vad Mathiesen (Aalborg University), Niclas Bentsen (Copen-hagen University) and Lorie Hamelin (Southern Denmark Univer-

sity). Financial support for this study was provided by a researchgrant (‘‘CEESA’’ 2104-06-0007) from the Danish Research Councilas well as from the Technical University of Denmark. The authorswish to thank two anonymous reviewers for detailed and valuablecomments.

Appendix A. Supplementary material

Additional information on inventory data for: biomass re-sources, energy conversion technologies, land use changes andmanagement of agricultural and biomass conversion residualscan be found in the Supporting Information (SI). Supplementarydata associated with this article can be found, in the online version,at http://dx.doi.org/10.1016/j.apenergy.2012.03.006 .

References

[1] Danish Energy Agency (DEA). Energistatistik 2009, Copenhagen (Denmark):Energistyrelsen (Danish Energy Agency); 2010.

[2] Danish Ministry of Climate. Energy and Buildings, Our future energy.Copenhagen (Denmark): Danish Ministry of Climate, Energy and Buildings;2011.

[3] Li YP, Huang GH, Chen X. Planning regional energy system in association withgreenhouse gas mitigation under uncertainty. Appl Energy 2011;88:599–611.

[4] Mathiesen BV, Lund H, Karlsson K. 100% Renewable energy systems, climatemitigation and economic growth. Appl Energy 2011;88:488–501.

[5] Mathiesen BV, Lund H, Karlsson K. IDAs Klimaplan 2050, baggrundsrapport –Tekniske systemanalyser, brændselsforbrug, drivhusgasser, samfundsøkonomiske konsekvenser, erhvervspotentialer, beskæftigelseseffekter samthelbredsomkostninger (IDAs Climate Plan 2050, background report in Danishand English), Copenhagen (Denmark): Danish Society of Engineers (IDA,Ingeniørforeningen Danmark); 2009.

[6] Mathiesen BV, Lund H, Hvelplund FK, Connolly D, Bentsen NS, Tonini D, et al.Hansen, CEESA 100% Renewable Energy Scenarios towards 2050. Aalborg(Denmark): Aalborg University; 2011. < http://www.ceesa.plan.aau.dk >(accessed 12.11).

Table 7

Results of the sensitivity analysis expressed as net differences ‘ D ’ compared with thebaseline result as presented in Fig. 4. Units are D Gg CO2-eq/PJ PES (GW), 10 6 D m 2/PJ(AC), D Mg N/PJ

PES (EP (N)), D Mg P/PJ

PES (EP (P)), and 10 6 D m 2/PJ

PES). The sensitivity

analysis for RME (2) and biochar (4) only affected the GW category ( y: yield; g :efciency; " : increase; ; : decrease).

Category Parameter 2050CSV 2050RME 2050BtL

GW (1) Willow ( y" / y; ) 1/+2 1/+2 2/+4(2) RME (iLUC ; ) – 11 –(3) BtL (g " ) – – 1(4) Biochar ( ; ) +2 +2 +4

AC (1) Willow ( y" / y; ) 1.6/+4 1.6/+4 5/+7(3) BtL (g " ) – – 2

EP (N) (1) Willow ( y" / y; ) 1/+2.5 1/+2.5 3/+4.5(3) BtL (g " ) – – 1

EP (P) (1) Willow ( y" / y; ) 0.1/+0.2 0.1/+0.2 0.2/+0.5(3) BtL (g " ) – – 0.1

LO (1) Willow ( y" / y; ) 60/+440 60/+440 445/+630(3) BtL (g " ) – – 100




[7] Franco A, Salza P. Strategies for optimal penetration of intermittent renewablesin complex energy systems based on techno-operational objectives. RenewEnergy 2011;36:743–53.

[8] Connolly D, Lund H, Mathiesen BV, Leahy M. The rst step towards a 100%renewable energy-system for Ireland. Appl Energy 2011;88:502–7.

[9] Cosmi C, Di Leo S, Loperte S, Macchiato M, Pietrapertosa F, Salvia M, et al. Amodel for representing the Italian energy system: the NEEDS-TIMESexperience. Renew Sustain Energy Rev 2009;13:763–76.

[10] Treffers DJ, Faaij APC, Spakman J, Seebregts A. Exploring the possibilities forsetting up sustainable energy systems for the long term: two visions for theDutch energy system in 2050. Energy Policy 2005;33:1723–43.

[11] Rubio Rodriguez MA, Ruyck JD, Diaz PR, Verma VK, Bram S. An LCA basedindicator for evaluation of alternative energy routes. Appl Energy2011;88:630–5.

[12] Tan RR, Aviso KB, Barilea IU, Culaba AB, Cruz Jr JB. A fuzzy multi-regionalinput–output optimization model for biomass production and trade underresource and footprint constraints. Appl Energy 2012;90:154–60.

[13] Mathiesen BV, Lund H, Noergaard P. Integrated transport and renewableenergy systems. Utilities Policy 2008;16:107–16.

[14] Lund H, Kempton W. Integration of renewable energy into the transport andelectricity sectors through V2G. Energy Policy 2008;36:3578–87.

[15] Wenzel H, Hauschild MZ, Alting L. Environ Assess Products 1997;1.[16] Lund H, Mathiesen BV. Energy system analysis of 100% renewable energy

systems: the case of Denmark in years 2030 and 2050. Energy

2009;34:524–31.[17] Hauschild M, Potting J. Spatial differentiation in life cycle impact assessment –the EDIP2003 methodology, Environmental News No. 80. Copenhagen(Denmark): Danish Ministry of the Environment; 2005.

[18] Jolliet O, Margni M, Charles R, Humbert S, Payet J, Rebitzer G, et al. IMPACT2002+: a new life cycle impact assessment methodology. Int J Life Cycle Assess2003;8:324–30.

[19] Finnveden GA, Hauschild MZ, Ekvall T, Guinee J, Heijungs JBR, Hellweg S, et al.Recent developments in life cycle assessment. J Environ Manage2009;91:1–21.

[20] Pre’ Consultants, Simapro 7.1, www. pre.nl. (accessed January 2009); 2009.[21] Cencic O, Rechberger H. Material ow analysis with software STAN. J Environ

Manage 2008;18(1):3–7.[22] Danish Energy Agency (DEA). Energistatistik 2008. Copenhagen (Denmark):

Energistyrelsen (Danish Energy Agency); 2009.[23] Klimatkommisionen (Danish Commission on Climate Change Policy). Green

energy – the road to a Danish energy system without fossil fuels. Copenhagen(Denmark): Danish Commission on Climate Change Policy; 2010.

[24] Callesen I, Grohnheit PE, Oesterga H. Optimization of bioenergy yield fromcultivated land in Denmark. Biomass Bioenergy 2010;34:1348–62.

[25] Boehmel C, Lewandowski I, Claupein W. Comparing annual and perennial

energy cropping systems with different management intensities. Agric Syst2008;96:224–36.[26] Lund H. Renewable energy strategies for sustainable development. Energy

2007;32:912–9.[27] Jungbluth N, Chudacoff M, Dauriat A, Dinkel F, Doka G, Faist EM, et al. Life cycle

inventories of bioenergy. Uster (Switzerland): Swiss Centre for Life CycleInventories, ESU-services, Ecoinvent report No 17; 2007.

[28] Hamelin L, Joergensen U, Petersen BM, Olesen JE, Wenzel H. Modelling theenvironmental consequences of direct land use changes from energy crops inDenmark: a consequential life cycle inventory, GCB Bioenergy; 2012.

[29] Schmidt JH. System delimitation in agricultural consequential LCA – outline of methodology and illustrative case study of wheat in Denmark. The Int J LifeCycle Assess 2008;13:350–64.

[30] Schmidt JH. Life cycle assessment of rapeseed oil and palm oil, PhD thesis.Aalborg (Denmark): Aalborg University; 2007.

[31] Cherubini F, Ulgiati S. Crop residues as raw materials for biorenery systems –a LCA case study. Appl Energy 2010;87:47–57.

[32] Hamelin L, Wesnaes M, Wenzel H, Petersen BM. Environmental consequencesof future biogas technologies based on separated slurry. Environ Sci Technol2011;45:5869–77.

[33] Hamelin L, Wesnæs M, Wenzel H, Petersen BM. Life cycle assessment of biogasfrom separated slurry, Environmental project 1329. Copenhagen (Denmark):Danish Environmental Protection Agency; 2010. < http://www.mst.dk/Publika-tioner/Publications/2010/07/978-87-92668-03-5.htm > (accessed 01.11).

[34] Roberts KG, Gloy BA, Joseph S, Scott NR, Lehmann J. Life cycle assessment of biochar systems: estimating the energetic, economic, and climate changepotential. Environ Sci Technol 2010;44:827–33.

[35] International Energy Agency (IEA). CO 2 emissions from fuel combustionhighlights. Paris (France): International Energy Agency (IEA); 2011.

[36] Zhu L, Zhang W, Liu W, Huang Z. Experimental study on particulate and NO xemissions of a diesel engine fueled with ultra low sulfur diesel, RME-dieselblends and PME-diesel blends. Sci Total Environ 2010;408:1050–8.

[37] Sun J, Caton JA, Jacobs TJ. Oxides of nitrogen emissions from biodiesel-fuelleddiesel engines. Progress Energy Combust Sci 2010;36:677–95.

[38] Mazzoleni C, Kuhns HD, Moosmuller H, Witt J, Nussbaum NJ, Chang MCO, et al.A case study of real-world tailpipe emissions for school buses using a 20%biodiesel blend. Sci Total Environ 2007;385:146–59.

[39] Wang WG, Lyons DW, Clark NN, Gautam M, Norton PM. Emissions from nineheavy trucks fueled by diesel and biodiesel blend without engine modication.Environ Sci Technol 2000;34:933–9.

[40] Abrahamson LP, Volk TA, Kopp RF, White EH, Ballard JL, editors. Willowbiomass producer’s handbook. Syracuse (NY): Short-Rotation Woody CropsProgram, SUNY-ESF; 2002.

[41] Ros J, Overmars K, Notenboom J. How to deal with indirect land-use change inthe EU Renewable Energy Directive? Den Haag/Bilthoven (The Netherlands):Netherlands Environmental Assessment Agency; 2010.

[42] Croezen HJ, Bergsma GC, Otten MBJ, Van Valkengoed MPJ. Biofuels: indirectland use change and climate impact. Delft (The Netherlands): CE Delft;2010.

[43] Kunikowski G, Gan ´ko E, Pisarek M, Nilsson LJ, Ericsson K, Witt J, et al. Review –regional energy crops experience. Warsaw (Poland): EC Baltic Renewable

Energy Centre ECBREC/IBMER; 2007.[44] Danish Energy Agency (DEA). Fjernkøling i Danmark (District Cooling inDenmark). Copenhagen (Denmark): Energistyrelsen (Danish Energy Agency);2007.

[45] Tommerup H, Svendsen S. Energy savings in Danish residential building stock.Energy Build 2006;38:618–26.

[46] Wittchen K. Potentielle energibesparelser i det eksisterende byggeriSkærpedekrav til nybyggeriet 2010 og fremover–Økonomisk analyse(Increased demandsfor buildings standards from 2010). Hørsholm(Denmark): Statens Byggeforskningsinstitut (Danish Building ResearchInstitute); 2009.

[47] Aggerholm S. Skærpede krav til nybyggeriet 2010 og fremover–Økonomiskanalyse (Increased demands for buildings standards from 2010). Hørsholm(Denmark): Statens Byggeforskningsinstitut (Danish Building ResearchInstitute); 2009.

[48] Johansen M, Hansen M, Mikkelsen J. Energibesparelser i erhvervslivet (Energysavings within industry and service). Copenhagen (Denmark): Dansk EnergiAnalyse A/S og Weel og Sandvig Aps; 2009.

[49] Lund H, Mathiesen B. Ingeniørforeningens Energiplan 2030 – Tekniskeenergisystemanalyser, samfundsøkonomisk konsekvensvurdering ogkvanticering af erhvervspotentialer. Baggrundsrapport (Danish Society of

Engineers’ Energy Plan 2030). Copenhagen (Denmark): Danish Society of Engineers (Ingeniørforeningen Danmark); 2006.[50] Nielsen O, Landex A, Rørbech J. Fremtidsscenarier for transport i Danmark

(Future scenarios for transport in Denmark). Copenhagen (Denmark):Ugebladet Ingeniøren; 2006.

[51] Ahrenfeldt J, Henriksen U, Jensen TK, Goebel B, Wiese L, Kather A, et al.Validation of a continuous combined heat and power (CHP) operation of atwo-stage biomass gasier. Energy Fuels 2006;20:2672–80.

[52] Arena U, Gregorio FD, Santonastasi M. A techno-economic comparisonbetween two design congurations for a small scale, biomass-to-energygasication based system. Chem Eng J 2010;162:590.

[53] Arena U, Zaccariello L, Mastellone ML. Fluidized bed gasication of waste-derived fuels. Waste Manage 2010;30:1212–9.

[54] Bentzen JD, Hindsgaul C, Henriksen UB, Soerensen LH. Straw Gasication in aTwo-Stage Gasier. In: Proceedings of the 12. European biomass conference;2002. p. 577–80.

[55] Fruergaard T, Astrup T, Ekvall T. Energy use and recovery in wastemanagement and implications for accounting of greenhouse gases andglobal warming contributions. Waste Manage Res 2009;27:724–37.

[56] Jungbluth N, Frischknecht R, Emmenegger MF, Steiner R, Tuchschmid M. Lifecycle assessment of BTL-fuel production: inventory analysis, RENEW,renewable fuels for advanced powertrains; ESU-services Ltd Del-5.2.7-inventory-nal.doc; 2007.

[57] Larsen J, Østergaard Petersen M, Thirup L, Wen Li H, Krogh Iversen F. The IBUSprocess – lignocellulosic bioethanol close to a commercial reality. Chem EngTechnol 2008;31:765–72.




II

Bioenergy Production from PerennialEnergy Crops: a Consequential LCA of 12Bioenergy Scenarios including Land Use

Changes

Tonini, D., Hamelin, L., Wenzel, H., Astrup, T.

Environmental Science and Technology, 46(24), 13521-13530



Bioenergy production from perennial energy crops: a consequentialLCA of 12 bioenergy scenarios including land use changesTonini Davide,† , * Hamelin Lorie,‡ Wenzel Henrik,‡ and Astrup Thomas†

† Department of Environmental Engineering, Technical University of Denmark, Miljoevej 115, 2800 Kgs. Lyngby, Denmark ‡ Institute of Chemical Engineering, Biotechnology and Environmental Technology, Faculty of Engineering, University of SouthernDenmark, Campusvej 55, 5230 Odense M., Denmark

*S Supporting Information

ABSTRACT: In the endeavor of optimizing the sustainability

of bioenergy production in Denmark, this consequential lifecycle assessment (LCA) evaluated the environmental impactsassociated with the production of heat and electricity from onehectare of Danish arable land cultivated with three perennialcrops: ryegrass ( Lolium perenne), willow (Salix viminalis) and Miscanthus giganteus. For each, four conversion pathways wereassessed against a fossil fuel reference: (I) anaerobic co-digestion with manure, (II) gasication, (III) combustion insmall-to-medium scale biomass combined heat and power(CHP) plants and IV) co-ring in large scale coal-red CHPplants. Soil carbon changes, direct and indirect land usechanges as well as uncertainty analysis (sensitivity, Mon-teCarlo) were included in the LCA. Results showed that global warming was the bottleneck impact, where only two scenarios, namely willow and Miscanthus co- ring, allowed for animprovement as compared with the reference (− 82 and− 45 t CO2-eq. ha− 1 , respectively). The indirect land use changes impact

was quantied as 310 ± 170 t CO2-eq. ha− 1

, representing a paramount average of 41% of the induced greenhouse gas emissions.The uncertainty analysis conrmed the results robustness and highlighted the indirect land use changes uncertainty as the only uncertainty that can signicantly change the outcome of the LCA results.

1. INTRODUCTIONThe ambition of the energy policy in Denmark is to reach a100% renewable energy system by 2050.1 Several studies have been conducted to design and optimize such a system, andthese all highlight the indispensability of a biomass potential of around 35− 50% of the overall energy consumption.2− 5 Thereare several reasons explaining why biomass is so attractive forenergy systems entirely free of fossil energy.6 Its key advantage,however, lies in the fact that it is storable, entitling it to be usedfor balancing the uctuating energy production fromintermittent sources like wind and solar power.1,2,6,7

Though biomass is a renewable energy source, it is notunlimited in supply, and does involve considerable environ-mental costs. One of the most critical costs of bioenergy relatesto its incidence on land use changes (LUC),8− 10 that is, theconversion of land from one use (e.g., forest, grassland or food/feed crop cultivation) to another use (e.g., energy cropcultivation).

One way to minimize these LUC impacts could be throughfavoring the cultivation of perennial energy crops (e.g.,perennial ryegrass, willow and Miscanthus) instead of annualcrops (e.g., maize, barley, wheat, sugar beet). In fact, it isacknowledged that perennial energy crops nowadays representthe most efficient and sustainable feedstock available for

bioenergy production in temperate regions.11 − 13 Amongothers, perennial energy crops generally present a moreefficient nutrient use than their annual counterpart, whichinvolves lower requirements for annual inputs of fertilizers, andconsequently lower environmental impacts related to fertiliza-tion.14 Moreover, in contrast to annual crops whose cultivationtends to accelerate the depletion of soil organic carbon (SOC),perennial energy crops allow for an accumulation of SOC.14

They generally also present higher yields, involve less soildisturbances due to their longer life cycle duration, and have a better incidence on biodiversity.12 For these reasons, this study focuses on bioenergy production from perennial energy cropsonly.

The goal of this study is to assess the environmental impactsassociated with the production of bioenergy (heat andelectricity) from 1 ha (10,000 m2) of Danish arable landcultivated with ryegrass, willow and Miscanthus , consideringfour diff erent biomass-to-energy (BtE) conversion pathways:(i) anaerobic co-digestion with manure, (ii) gasication, (iii)

Received: June 20, 2012Revised: October 8, 2012 Accepted: November 5, 2012

Article

pubs.acs.org/est

© XXXX American Chemical Society A dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXX



combustion in small-to-medium scale biomass combined heatand power (CHP) plants and iv) co-ring in large scale coal-red CHP plants.

2. MATERIALS AND METHODS2.1. Life Cycle Assessment Model. 2.1.1. Scope and

Functional Unit. The environmental assessment presented inthis study was performed using consequential life cycleassessment (LCA).15,16 The functional unit upon which allinput and output ows were expressed was 1 ha of agriculturalland used to grow the selected energy crops. The geographicalscope considered for the LCA was Denmark, that is, the datainventory for crops cultivation and BtE plants were specic forDanish conditions. Similarly, the legislative context of Denmark (e.g., fertilization) was considered. The temporal scopeconsidered was 20 years, i.e., all assessed systems were operatedfor 20y duration.

2.1.2. Impact Assessment. The life cycle impact assessment was carr ied out according to the Danish EDIP 2003method17,18 for the environmental impact categories global warming (aggregated emissions over a 100 years horizon)(GW) and aquatic eutrophication (distinguishing betweennitrogen and phosphorus being the limiting nutrient forgrowth) (EP (N) and EP (P), respectively). To this, an impactcategory named “ Phosphorous as resource” was added in orderto re ect the benets associated with phosphorus (P) savings, based on the Impact 2002+ method.19 Background LCA data were based on the Ecoinvent v.2.2 database, and the assessment was facilitated by the LCA software SimaPro 7.3.3.20 Fore-ground LCA data essentially included Danish-specic data foragricultural and energy conversion processes, and the impacts

associated with capital goods (foreground data only) as well asthose related to transportation of the residues (i.e., ash anddigestate) have been excluded.

2.2. Scenarios Modeling and System Boundary. Thesystems assessed considered three perennial crops (ryegrass, willow and Miscanthus) and four BtE conversion technologies(anaerobic co-digestion, gasication, combustion in small-to-medium scale biomass CHP plants and co-ring in large scalecoal- red CHP plants). A total of 12 scenarios have therefore been assessed. The system boundary conditions are illustratedin Figure 1, for the case of ryegrass anaerobic co-digestion. Theprocess ow diagrams for the other scenarios are similar,though the pre-treatments and the ows diff er, as shown inTable S2 and Figures S1− S11 of the Supporting Information(SI).

For all BtE technologies, the energy produced was

considered to be used for CHP production, thereby substituting the production of marginal heat and power. Inthe present study, the marginal electricity source was assumedto be from coal-red power plants conformingly with refs ,22and the marginal heat from natural gas-based domestic boiler,this being the fuel which is most likely to react to a marginalchange in the heat demanded/supplied23 (further detailed inthe SI).

As illustrated in Figure 1, the digestate produced fromanaerobic digestion was used as a fertilizer (for N, P, and K), which avoided marginal mineral N, P, and K fertilizers to beproduced and used, based on the content of N, P, and K of thedigestate. The marginal N, P, and K fertilizers considered werecalcium ammonium nitrate, diammonium phosphate, andpotassium chloride, respectively, conformingly with refs 14

Figure 1. Process ow diagram for the bioenergy scenario anaerobic co-digestion of ryegrass with raw pig manure. Electricity and heat producedrepresent net values (i.e., plants own consumptions have been subtracted). (* ) Not all the converted land is to be cultivated in barley, and not all theDanish barley displaced is replaced, due to various market mechanisms. Values rounded (2 signicant digits).

Environmental Science & Technology Article

dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXXB



and 24. Further, based on the model from ref 24, it wasconsidered that the manure portion used for co-digestion would have otherwise been stored and applied on land, withoutdigestion or other treatment.The three thermal bioenergy scenarios (i.e., gasication,combustion, and co-ring) implied negligible residual uncon- verted carbon that is found in the bottom ashes, y ashes, andeventual wastewater. The bottom ashes were assumed to beused for road construction, substituting for natural aggregates, whereas the y ashes were assumed to be utilized for back llingof old salt mines with negligible environmental impacts.25

Treatment of wastewater was not included. All bioenergy scenarios involved the use of Danish

agricultural land in order to grow the energy crops. In acountry like Denmark, where 68% of the total land is used forcropland and where policies have been adopted in order todouble the forested area (nowadays representing ca. 13% of thetotal land),26 very limited conversion from forest or alike naturetypes is occurring. Most likely, the land needed to grow theenergy crops will be taken from actual Danish cropland,involving that one crop cultivated today will be displaced. Sucha displaced crop is, in consequential LCA, referred to as themarginal crop. In this study, the marginal crop was assumed to be spring barley, based on.22,27,28 Based on the consequentialLCA logic, as well as on recent studies,9,29,30 this resulting dropin supply of Danish spring barley will cause a relative increase inagricultural prices, which then provide incentives to increasethe production elsewhere. Such increased crop production may stem from both increased yield and land conversion tocropland, the latter being also referred to as indirect land usechange (iLUC).9,29,30 As illustrated in Figure 1, and as in recentiLUC studies,10,31,32 this study included the environmentalimpacts of the latter only.

2.3. Life Cycle Inventory (LCI). 2.3.1. Crops. The LCI of all crops was based on a recent Danish consequential LCI,14

which comprises all processes involved during the cultivationstage, up to harvest. This included the tillage activities, liming,propagation (seed, rhizome, and cutting production), plantprotection, fertilization, sowing/planting, harvest, and transportfrom farm to eld. A sandy loam soil has been considered for allcrops, as well as precipitations of 964 mm y − 1. For Miscanthusand willow, the C turnover rate in the topsoil was considered to be reduced by 25% in response to the absence of tillage overmany years. For all crops, the fertilization operations wereperformed in conformity with Danish regulations,33,34 involvingan upper limit for the amount of N to be applied on the eld, both as mineral fertilizer and animal slurry.

Based on ref 14, the life cycle considered for perennialryegrass (short-term ley), willow and Miscanthus plantations were respectively 2 years, 21 years (6 cuts; 3 years harvest cycle, but rst harvest after 4 years; 1 year establishment; 1 yearpreparation before planting), and 20 years (18 cuts; 1 yearestablishment; 1 year preparation before planting). Given the20 year temporal scope of the LCA, this means that the lifecycle of ryegrass, willow and Miscanthus is respectively occurring 10, 0.95, and 1 time. Further, it was consideredthat ryegrass was harvested in summer, willow in the vegetativerest period (in the period around November to February) and Miscanthus during the spring season.

2.3.2. BtE Conversion Technologies and Pre-treatments. Anaerobic digestion was modeled as mesophilic co-digestion of the respective energy crops with raw pig manure. Manurerepresents one of the most abundant domestically available

biomass resources in Denmark (ca. 23− 34 PJ), which isnowadays signicantly underexploited for energy production.5

The current management of raw manure consists to store it inan outdoor structure until it can be used as an organic fertilizeron agricultural land, which leads to large impacts on mostenvironmental compartments, mainly global warming andeutrophication.24 Hence, co-digestion of manure with carbon-rich biomass may represent a viable alternative to produce bioenergy and improve manure management. The modeledmethane yields for ryegrass, willow, Miscanthus and raw pigmanure were, respectively, 290, 240, 250, and 320 N-m3 t− 1 VS(see SI). Based on ref 24 the mixture of crop and raw pigmanure was calculated in order to ensure a biomass mixtureinput having a dry matter (DM) content of 10% after the rstdigestion step. The resulting ratio manure:crop (fresh weight basis) for co-digestion of ryegrass, willow and Miscanthusequaled 5.7, 6.4, and 6.7, yielding respectively 140, 160, and 130MJ CH4 ha− 1 (SI Table S9). Consumption of electricity (2% of the energy in the biogas) and heat (to heat up the substratesfrom 8 to 37 °C) was modeled according to.24 Fugitive CH4emissions were taken as 1% of the produced CH4 , based onrecent studies.24,35,36 More details on the modeling of anaerobicdigestion can be found in the SI.

Gasi cation was modeled as uidized bed gasication basedon a number of reviewed studies (SI Table S5). The resultingcold gas and carbon conversion efficiency (CGE and CCE) was70% (± 15%) and 95% (± 4%), respectively. Consumption of electricity (26 kWh t− 1 DM) was based on ref 36.

Combustion was modeled as direct biomass combustion insmall-to-medium scale biomass CHP plants, based on athorough review of (mainly Danish) biomass CHP plants (SITable S6). Average net electricity and heat efficienciesinventoried from this review were 27% (± 2%) and 63%(± 7%), respectively. Co-ring in large scale coal-red CHPplants was likewise modeled, resulting to net electricity andheat efficiencies of 38% (± 3%) and 52% (± 8%), respectively (SI).

The air emissions from biogas and syngas combustion in gasengines as well as from biomass combustion in CHP plants were based on ref 37 (SI Table S7). Both biogas and syngas were assumed utilized in a gas engine with an average grosselectricity and heat efficiency of 38% (± 4%) and 52% (± 8%)(relative to the LHV of the input-gas).

Pre-treatments included on eld drying (ryegrass, for all BtEconversion technologies) and natural drying (willow, forgasi cation, direct combustion and co-ring), size comminution(all crops, for all BtE conversion technologies except directcombustion) as well as steam pre-treatment for breaking thelignocellulosic structures of Miscanthus and willow undergoinganaerobic digestion. All these pre-treatments are furtherdetailed in the SI.

2.3.3. Other Processes. Additional processes modeled in theLCA were: crops and digestate storage, use on land (UOL) of the digestate, treatment of residues from thermal BtEtechnologies and transportation. A detailed description of these processes can be found in the SI.

2.4. Carbon and Nitrogen Flow Analysis. Carbon andnitrogen ows are two of the most important ows responsiblefor the environmental impacts involved in bioenergy systems.Therefore, the C and N ows of all the scenarios assessed inthis study have been disaggregated and calculated for all themajor processes involved. This included the soil C changesresulting from the cultivation stage, which were calculated with


dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXXC



the dynamic soil C model C-TOOL,38,39 as detailed in ref 14for all crop systems. The modeling of the other C and N ows was based on the equations listed in the SI. The carbon andnitrogen ow analysis was facilitated by the software STAN

40

allowing a quantication of the uncertainties for the mostsensitive parameters (SI Table S17) and to reconcile the data when necessary.

2.5. Direct and Indirect Land Use Changes Impacts. Asearlier explained, the LCA system established in this study considers that the land used for cultivating the energy crops would have otherwise been used for cultivating spring barley (with straw incorporation) for the food/feed market (Figure 1).The direct land use change (dLUC) consequence of thistranslates into the environmental impacts of cultivating theselected energy crops instead of spring barley (Figure 1). Theenvironmental impacts from spring barley cultivation have beenincluded on the basis of the data from ref 14.

The iLUC consequence corresponds to the environmentalimpact of converting land nowadays not used for cropcultivation to cropland, as a result of the induced demand forthe displaced spring barley. To quantify this impact, it isnecessary to identify (i) how much land is converted and where; and (ii) which types of land are converted (biometypes). So far, most studies attempting to quantify themagnitude of iLUC used econometric models to this end, forexample, refs 9,10,29,31, and 32 where the economic and biophysical/agricultural systems are combined into one singlemodeling framework. A comprehensive overview of partial andgeneral equilibrium models that can be used to model iLUC isgiven in ref 41.

Most of available iLUC studies to date focused on biofuelmandates for a variety of shock sizes, and as such are difficult to be used directly for other applications. In,29 however, the iLUCconsequences in terms of points (i) and (ii) above areidenti ed, for a marginal increase in wheat consumption in fourdiff erent countries, including Denmark. This was done using amodi ed version of the general equilibrium GTAP model.42 Inthe present study, the results of ref 29 for Denmark have beenused as a proxy to estimate how much land is converted (due tothe increased spring barley demand) and where. However, theCO2 impact of land conversion is not estimated in.29 In orderto do so, the soil and vegetation C data from the Woods HoleResearch Centre, as published in,9 have been used, and the CO2emitted due to land conversion was calculated based on themethodology published in.43 Based on this methodology, it wasconsidered that 25% of the C in the soil was converted to CO2for all types of land use conversion, except when forests wereconverted to grassland, where 0% was converted. Further, it wasconsidered that 100% of the C in vegetation was converted toCO2 for all forest types as well as for tropical grasslandconversions, while 0% was converted for the remaining biometypes (e.g., shrub land, non-tropical grassland, chaparral).

2.6. Sensitivity and Uncertainty Analysis. Two types of uncertainties were addressed in this study (for the GW impactonly), namely scenario and parameter uncertainties. While theformer deals with the uncertainty due to the intrinsic modelingchoices (in terms of system boundary and marginaltechnologies/products), the latter covers the uncertainty related to the quantication of the values used in the LCA model.

Parameter uncertainties were addressed through a Mon-teCarlo analysis (number of simulations: 1000), whereasscenario uncertainties were addressed through sensitivity

analyses. These included (a) variation (min-max) of theiLUC impacts with respect to CO2 emissions (vs mean valueassumed as baseline); (b) winter wheat as the marginal crop forDenmark (vs spring barley as baseline); (c) coal-based heatproduction as the marginal technology for heat generation (vsnatural gas-based as baseline); (d) natural gas power plant asthe marginal technology for electricity generation (vscondensing coal power plant as baseline); (e) mono-digestionof the crops (vs co-digestion with manure as baseline); (f) pre-treatment of pelletization before co-ring (vs “ no pelletization”

as baseline). Each of these changes was tested individually toassess the inuence of the individual change on the overall LCA results.

A thorough description of the methodology used forsensitivity and uncertainty analysis can be found in the SI.

3. RESULTS AND DISCUSSION3.1. Carbon and Nitrogen Flows. The induced C and N

ows for ryegrass, willow and Miscanthus are presented inFigures S13− S18 (SI).

As illustrated in SI Figures S13− S15, more than 85% of theC input to the energy crop system (the most notable being theuptake from the atmosphere) ends up emitted as CO2 , whetheras a result of the cultivation stage or as a result of the nalenergy use. As indicated in refs 8,44, many bioenergy studiesreport rather diff erent results, as the biogenic CO2 emissionsfrom the cultivation stage (releases from manure and residuesnot entering the soil C pool), which here represents 50− 57%(SI Figures S13-S15) of the C input fate, are not accounted for.This highlights the importance of the error made if a completesystem-based mass balance, such as the one performed in thisstudy, is not considered.

The C from atmospheric uptake was similar for all the threecrops (about 11− 12 t C ha

−

1 y −

1): for all crops, only about half of this C ended up in the harvested biomass, the other half ending up in the non-harvested above- and below-groundresidues (SI Figures S13− S15). The biogenic CO2 emissionrelated to crop cultivation (6.1 to 6.9 t CO2

− C ha− 1 y − 1) wasalso in the same order of magnitude for all crops (SI FigureS13− S15; Table S8). The biogenic carbon emission from thenal energy use, however, varied signicantly more (2.9 to 6.0 t

CO2− C ha− 1 y − 1), as detailed in SI Table S8. This reects the

importance of two main parameters: the crop yield and the BtEtechnology. In fact, the biogenic CO2 emission from the nalenergy use was the greatest for thermal treatments (combustionand gasication), where 95− 100% of the carbon was emitted asCO2 , whereas it was signicantly lower for biological treatment(anaerobic co-digestion), where only ca. 40− 46% of the crop(and raw manure) carbon was gasied (SI Table S8). Thisunconverted C during anaerobic co-digestion is ultimately applied on land, through the digestate. However, this did notrepresent a signicant carbon sink, as more than two-thirds of this C was released as CO2 , rather than sequestrated in the soil(SI Figure S13− S15). This is in accordance with previousndings (e.g., ref 24).The variation in SOC due to dLUC was positive (i.e., the

SOC content was increased) for all crop systems. This wasexpected, since spring barley, an annual crop with a much lower yield than any of the perennial energy crops considered here,involves losses instead of gains in soil C, as illustrated in.14 Themodeled Δ SOC was very similar for the three crops (about 0.7t C ha− 1 y − 1). The avoided CO2 emissions resulting from thesubstitution of fossil carbon were proportional to the amount of


dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXXD



bioenergy produced; this ranged from 3.9 (anaerobic co-digestion of Miscanthus with raw pig manure) to 8.3 (co-ringof willow) t C ha− 1 y − 1 (SI Table S8).

As opposed to C, the outputs of N ows were morediversied among the individual ows. The most signicant Nows occurred during the UOL of the digestate for the

anaerobic co-digestion scenarios, and during the cultivationstage for the other scenarios (SI Figures S16− S18; Table S8).Ryegrass showed the highest emissions of N during thecultivation phase; these occurred as a consequence of thehigher nitrogen fertilizer requirements of ryegrass (450 kg Nha

−

1 y −

1) compared with willow (170 kg N ha−

1 y −

1) and Miscanthus (71 kg N ha− 1 y − 1). These fertilization rates (and therelated N-based emissions) are based on today ’ s practices, butshould be seen as reecting the highest end of the interval. Infact, Miscanthus and willow are relatively new crops, and it can be expected that lower application rates will be required asinsight is gained on the optimal management of thesecrops.45,46 Similarly, lower N application could be consideredfor ryegrass dedicated to bioenergy, where protein productionis not the focus (as in the case of forage ryegrass). The N-related emissions at the UOL stage (anaerobic co-digestionscenarios) were similar for all the three crops, as a consequenceof the Danish legislation for fertilization xing the maximalamount of N to be applied in agricultural elds.33,34 Overall,

NO3− and NH3 emissions were the most signicant N-

emissions.3.2. Indirect Land Use Changes. The iLUC impacts of

the studied bioenergy systems were the same for all scenarios(Figure 2a), as they all had the same “ point of origin” : theconversion of 1 ha of Danish land (cultivated with spring barley) to energy crops. As shown in Table 1 (and furtherdetailed in the SI), these iLUC impacts were estimated to 310 tCO2-eq. ha− 1 (± 170 t CO2-eq. ha− 1). The impacts wereannualized over a period of 20 years in accordance with IPCC47

and with prominent European legislation,48 corresponding to

about 16 t CO2-eq. ha− 1

y − 1

(or 70−

130 g CO2-eq. MJ− 1

y − 1

). Although currently debated and relatively uncertain,49 theiLUC impact quantied here can contribute with importantlearnings: (i) it is not zero; and (ii) it may cover a signicantproportion of the overall global warming impact (Figure 2a)(between one-third and half of the positive contributions,depending on the scenario), and cancels out the otherwiseavoided GHG emissions in the scenarios. Moreover, it should be highlighted that the 310 t CO2-eq. ha− 1 obtained here only covers the GHG related to the net expansion resulting from themodeling of ref 29 and does not include the GHG related tothe intensi cation of crop production (which accounts, basedon the results of (29), to about 30% of the displacementresponse). This suggests that the “ real” impact may actually behigher. The only other LCA study 50 the authors were aware of

Figure 2. LCA results for (a) global warming (over a 100 year horizon, t CO2-eq. ha− 1); (b) aquatic N-eutrophication (kg N ha− 1); c) aquatic P-eutrophication (kg P ha− 1); and d) phosphorus as resource (kg P ha− 1). All systems represent a 20 year time scope.


dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXXE



attempting to quantify iLUC on the basis of an hectare of landdisplaced (and not a biofuel mandate shock) led to aconsiderably higher value, that is, 440− 560 t CO2-eq. ha− 1

(considering a 20 years period and only conversion of forest). Although it cannot be directly compared, our annualized iLUC value (70− 130 g CO2-eq. MJ− 1 y − 1 , calculated dividing theannualized iLUC impact by the energy yielded by 1 hacultivated with the crops, dry basis) lies within the range of values found in ref 10 for marginal increases in the demand for biofuels.

In this study, the assessment of global warming was based onthe IPCC AR4 methodology,51 where GHG are summed up

over a dened time horizon, which in LCA is commonly takenas 100 years (as in this study). The use of this approach may however be seen as a limitation when emission releasesoccurring at diff erent times (e.g., year 0 and year 13) areinvolved, as these releases are then summed together despitethat their end points of analysis are diff erent (e.g., year 100 and year 113). In recent years, a number of studies have proposedmethodologies to address this aw, where many emphasizedthe particular case of iLUC (e.g., refs 43,52, and 53). As thesemethodologies are still in their early development stage, theglobal warming results presented in this study are based on theIPCC methodology. However, the importance of time-

Table 1. Estimation of the iLUC CO 2 Impact a

biomes convertedbtype of

conversionc regionc ,d m2 t− 1

wheatc ,φC in vegetation

(t ha− 1)eC in soil(t ha− 1)e

CO2− C lost

(t C t− 1 wheat) f CO2 lost (t CO2

t− 1 wheat)CO2 lost

(t CO2 ha− 1) g

savanna (taken as shrub land) 100%cropland

xss 140 ± 86 4,6 30 0.11 ± 0.06 0.39 ± 0.24 2.2 ± 1.3

African tropical evergreen forest(taken as tropical rain forest)

100%cropland

xss 140 ± 86 130 190 2.5 ± 1.5 9.1 ± 5.5 52 ± 31

open shrubland (taken as shrubland)

100%grassland

xss 81 ± 49 4,6 30 0.06 ± 0.04 0.22 ± 0.13 1.3 ± 0.8

temperate evergreen forest 100%cropland

xeu15 57 ± 34 160 130 1.1 ± 0.7 4.0 ± 2.4 23 ± 14

temperate deciduous forest 100%cropland

xeu15 57 ± 34 120 130 0.87 ± 0.52 3.2 ± 1.9 18 ± 11

dense shrub land (taken astemperate grassland)

46%cropland;54%grassland

xeu15 250 ± 148 7,0 190 1.2 ± 0.7 4.3 ± 2.6 24 ± 15

tropical evergreen forest 100%cropland

bra 180 ± 70 200 98 4.0 ± 1.6 15 ± 6 83 ± 33

savanna (taken as grassland) 100%

grassland

bra 41 ± 16 10 42 0.04 ± 0.02 0.16 ± 0.06 0.91 ± 0.36

grassland/steppe (taken astemperate grassland)

100%cropland

xsu 91 ± 55 10 190 0.43 ± 0.26 1.6 ± 0.9 9.0 ± 5.4

temperate evergreen forest 100%grassland

xsu 45 ± 27 160 130 0.88 ± 0.43 3.2 ± 1.6 18.3 ± 9.1

temperate deciduous forest 100%grassland

xsu 45 ± 27 140 130 0.76 ± 0.37 2.8 ± 1.3 16 ± 8

savanna (taken as tropicalgrassland)

100%cropland

aus 110 ± 64 18 42 0.31 ± 0.18 1.1 ± 0.7 6.4 ± 3.8

open shrubland & grassland/steppe (taken as tropicalgrassland)

100%grassland

aus 37 ± 22 18 42 0.11 ± 0.06 0.39 ± 0.23 2.2 ± 1.3

boreal deciduous forest (taken astemperate deciduous forest)

100%cropland

can 97 ± 58 140 130 1.6 ± 1.0 6.0 ± 3.6 34 ± 20

boreal evergreen forest (taken astemperate evergreen forest)

100%grassland

can 10 ± 6 160 130 0.16 ± 0.10 0.59 ± 0.35 3.3 ± 2.0

grassland/steppe (taken asgrassland)

100%cropland

xla 35 ± 21 10 42 0.04 ± 0.02 0.14 ± 0.08 0.77 ± 0.46

tropical evergreen forest 100%cropland

xla 35 ± 21 200 98 0.79 ± 0.48 2.9 ± 1.7 17 ± 10

savanna + dense shrub land(taken as grassland)

100%grassland

xla 16 ± 10 10 42 0.02 ± 0.01 0.063 ± 0.038 0.36 ± 0.22

open shrub land (taken aschaparral)

100%grassland

usa 68 ± 41 40 80 0.14 ± 0.08 0.50 ± 0.30 2.8 ± 1.7

total 1500 ± 880- 15 ± 8 54 ± 30 310 ± 170a Eventual inconsistencies due to rounding (numbers are reported with 2 signicant digits). bIndicated biomes are as in ref 29. When the biomesmentioned in ref 29 did not gure in the biomes from the Woods Hole Research Centre data,9 an equivalent was considered, which is indicated between parentheses, when it applies. cBased on the results from ref 29. d With xss: Sub-Saharan Africa, excluding Botswana, Lesotho, Namibia,South Africa and Swaziland; xeu15: EU-15, excluding Denmark; bra: Brazil; xsu: Former Soviet Union, excluding the Baltic States; aus: Australia;can: Canada; xla: South America, excluding Brazil and Peru; usa: United States. As indicated in ref 29 this aggregation covers 92% of the total netexpansion. eFrom the Woods Hole Research Centre, as published in ref 9. f Considering that 25% of the C in soil is converted, for all biomes, except when forest is converted to grassland, where 0% of soil C is converted; 100% of the C in vegetation is converted for all forest biomes; 100% of the Cin vegetation is converted for tropical grasslands; 0% of the C in vegetation is converted for all other biomes. g The conversion per ha is madeconsidering that it is 1 ha of spring barley that is initially displaced, with a yield of 4.9 t DM ha− 1 and a DM content of 85% of the crop fresh matter,

based on ref 14.φ

The maximal and minimal range are based on the qualitative description of the uncertainty related to the biomes conversion resultsmade by ref 29. The levels identied as “ very good” , “ good” and “ moderate” were considered as an uncertainty of ± 20%, 40%, and 60%, respectively.


dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXXF



dependency was assessed for the cultivation of Miscanthus(including iLUC), based on the methodology described in ref 53 (SI). This specic simulation indicated that accounting fortime-dependency would increase our GWP by ca. 40%. Suchincrease was also suggested by the results of ref 52 for adiff erent bioenergy case.

3.3. LCA Results. The environmental impacts related to the12 bioenergy scenarios assessed are shown in Figure 2 for theselected impact categories. Impacts/savings for the individual bioenergy scenarios were obtained by subtracting the avoidedimpacts (negative values in the gures) from the inducedimpacts (positive values). The zero axis represents thereference: any net value below the zero axis thus indicates anenvironmental improvement compared with the fossil fuelreference (in which: electricity and heat are provided by coaland natural gas, the hectare of land is used for spring barley cultivation, and manure is not digested).

On the selected impact categories, global warming appearscritical as only two scenarios indicate overall savings for thiscategory compared with the fossil fuel reference. Only co-ringof willow and Miscanthus indicated net overall savings, that is,these were the only two scenarios for which an environmental bene t, GHG-wise, was identied in relation to using 1 ha of land for bioenergy. However, the magnitude of the global warming impacts found in this study (between − 82 and 270 tCO2-eq. ha− 1 over 20 years) was much higher than previousresults from literature. For instance, ref 54 calculated a saving between ca.− 18 and− 35 t CO2-eq. ha− 1 y − 1 (ca. − 360 to− 700t CO2-eq. ha− 1 over 20 years) for bioenergy systems based on willow and Miscanthus plantations in Ireland;55 quanti edsavings about − 25 t CO2-eq. ha− 1 y − 1 (about − 500 t CO2-eq.ha− 1 in 20 years) for bioenergy systems based on Miscanthusplantations in Italy;56 estimated a saving between ca.− 10.4 and−

11.1 t CO2-eq. ha−

1 y −

1 (about −

210 to −

220 t CO2-eq. ha−

1in 20 years) for Miscanthus and willow plantations in the UK.The reason for these diff erences is that this study, as opposed tothe previous, considered iLUC, which has tremendoussigni cance on the overall GHG balance as earlier discussed.

As illustrated in Figure 2a, the 35% GHG emission savingrequired in the EU Renewable Energy Directive48 for biofuelsand bioliquids (as compared with the same energy providedfrom fossil fuels) has been used as a comparative measure of theGHG reductions achieved in the individual scenarios (althoughthe directive does not apply to these scenarios), see calculationdetails in the SI. As shown in Figure 2a, none of the assessed bioenergy scenarios would comply with a 35% GHG reductiontarget. This highlights the difficulties for bioenergy to compete with fossil fuels for producing heat and power. Though otherrenewable energy sources (e.g., wind, solar, hydro) should beprioritized, biomass (residual and energy crops) remainsneeded in a renewable energy system for its intrinsic versatility.2− 5 In this perspective and in the light of Figure 2a,co- ring or efficient combustion of willow and Miscanthus can be highlighted as preferable options for producing bioenergy from perennial crops, both in relation to global warming butalso to the other impact categories assessed (aquatic P and Neutrophication, P resource savings).

Co- ring and combustion provided the smallest global warming impacts for all crops. The environmental performanceof co- ring was directly related to the higher electricity efficiency of these plants (about 38% relative to the LHV of the fuel, wet basis), and consequently to the larger amount of marginal coal electricity substituted. Co-ring of willow

provided the largest savings, mostly because of the benecialdLUC, higher yield and minimal pre-treatment required.Similarly, the environmental performance of combustion wasdue to the high overall energy recovery as heat and electricity (about 90% relative to the LHV of the fuel, wet basis). Asopposed to combustion and co-ring, anaerobic co-digestionand gasication involved a conversion to gas before energy generation, thereby inducing additional losses (SI Table S9).Therefore, less electricity and heat were produced andsubstituted, resulting in larger net GW impacts from thesetechnologies. Further, UOL of the digestate contributed with aGW impact comparable to the one of iLUC, i.e., ranging between 280 ( Miscanthus) and 370 (willow) t CO2-eq. ha− 1 ,primarily connected to the release of biogenic carbon notentering the soil C pool (quantied in Figure S13− S15 of theSI). This cannot be directly visualized in Figure 2a, whichpresents the net impact of UOL (digestate minus raw manure).Co-digestion also resulted in GHG savings associated withavoiding raw manure management, which would otherwise bestored and applied on land without digestion.24 These savingsdepended on the amount of manure co-digested (per hectare),that is, the more manure co-digested (to meet the 10% DM inthe input-mixture), the larger the savings were. This alsoapplied to aquatic N-eutrophication, where the impacts weremuch higher for ryegrass because of the higher N content of thecrop.

Figure 2 highlights the signicance of dLUC for all scenariosand impact categories, where changing from spring barley toperennials generally resulted in environmental benets. Forglobal warming, this reects two main points. First, that theperennial crops considered in this study have a much greater Cuptake than spring barley. Second, that they are also moreefficient systems for converting the C uptake to useful C (i.e.,more C in the harvested biomass, less C in the residues,therefore less C lost as CO2 emissions during the cultivationstage). For the other impact categories, the dLUC results forryegrass diff ered from those of Miscanthus and willow. Figure2b for example reects the high load of N fertilizers applied inthe ryegrass system, which resulted in much higher N leachingthan in the reference (barley cultivation), while willow and Miscanthus systems resulted in a dLUC improvement. On theother hand, as half of the N fertilizers used during cultivationcame from animal slurries14 (which also contain P), no mineralP fertilizers needed to be applied for ryegrass, as opposed to allother crop systems, which explains the greater P savings for thiscrop in connection with dLUC (Figure 2d). It should however be kept in mind that the high N-leaching results for ryegrassshould be seen as a maximum, as ryegrass-for-bioenergy likely requires less N than ryegrass-for-fodder in order to reach thesame yields as considered in this study.

In Figure 2d, the category “ others” re ects the net induced Pfertilizers: since fertilization is by law based on crops N balance,33,34 even though anaerobic digestion allows fornutrients recycling, the higher nutrients content of theproduced digestate involves that relatively more P was appliedin excess in the co-digestion scenarios compared with thereference (use on land of raw pig manure), thus decreasing theoverall P-saving potential and increasing leaching (Figure 2c).P-leaching was less for willow as a consequence of the lower Pcontent of the crop.

The results of the sensitivity analyses highlighted that the variation of the iLUC impacts played the most important rolefor GW; with minimum iLUC impacts (Table 1) all bioenergy


dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXXG



scenarios for willow and Miscanthus as well as co-ring of ryegrass achieved environmental savings on GW (SI FigureS19). Co-ring and combustion of willow and Miscanthus evenreached the 35% GHG reduction target. In all other analyses,the individual changes in assumptions did not alter theconclusions relative to the baseline. However, the diff erentassumptions made regarding marginal energy and cropdecreased or increased the magnitude of the impacts or savingsin all scenarios (SI Figure S19). In the case of mono-digestion,GW impacts were signicantly increased as compared with theirlevels in the co-digestion scenarios (increase between 110 and160 t CO2-eq. ha− 1), re ecting the tremendous benetsobtained when avoiding conventional manure management.Co-digestion with manure shall therefore be favored in order tooptimize the GW savings associated with anaerobic digestion.The sensitivity analysis also demonstrated that additionalpelletization and milling of the biomass in the co-ringscenarios would decrease the GW performance of thesescenarios to a level very close to direct biomass combustion.The results of the MonteCarlo simulation for GW (SI TableS18) supported the ranking of the bioenergy scenarios found with the baseline scenarios, demonstrating that despite thesigni cant uncertainties, the results obtained were robust. Forgasi cation, combustion and co-ring, it also highlighted that it was not clear whether the willow scenarios really yielded greatersavings than the Miscanthus scenarios.

Overall, co-ring of Miscanthus and willow appeared to bethe options with the best environmental performance. It shouldhowever be realized that a main driver for future utilization of biomass may be to balance electricity generation from

uctuating energy sources, such as wind and solar power.Not all biomass combustion technologies may be suited forthis, especially when co-generation of heat is important as suchplants can have a xed production ratio between electricity andheat. Anaerobic digestion as well as gasication of biomass, onthe other hand, may be operated more exible without similarconstraints. Additionally, syngas or biogas off ers the exibility of storage. On this basis, improving the environmentalperformance of these BtE conversion technologies would bedesirable. For anaerobic digestion, a solution may be to favormanure-based biogas along with co-substrates not involvingiLUC (e.g., straw, organic municipal household waste, garden waste) as well as in boosting the digestion process by othermeans (e.g., digestion in series, enzymatic pre-treatment,addition of hydrogen, etc.).

■ ASSOCIATED CONTENT

*S

Supporting Information Additional information on marginal energy technologies andfertilizers, LCA process ow diagrams, LCI of crops and BtEconversion technologies, carbon and nitrogen ow charts,energy balance, GWP time-dependency, iLUC and modelingequations as well as sensitivity and uncertainty analyses. Thismaterial is available free of charge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATION

Corresponding Author*Phone: 0045 45251699. E-mail: [email protected] authors declare no competing nancial interest.

■ ACKNOWLEDGMENTSThe work presented in this paper is a result of the researchproject Coherent Energy and Environmental System Analysis(CEESA), partly nanced by The Danish Council for StrategicResearch.

■ REFERENCES(1) Lund, H.; Mathiesen, B. V. Energy system analysis of 100%

renewable energy systems: The case of Denmark in years 2030 and2050. Energy 2009 , 34 , 524− 531.(2) CEESA 100% Renewable Energy Scenarios Towards 2050; Aalborg

University: Aalborg, Denmark, 2011; http://www.ceesa.plan.aau.dk/digitalAssets/32/32603_ceesa_ nal_report_samlet_02112011.pdf (accessed April 30, 2012).(3) The Road to a Danish Energy System without Fossil Fuels. Summary

of the work, results and recommendations of the Danish Commissionon Climate Change Policy: Copenhagen, Denmark, 2010; http:// www.ens.dk/en-US/pol icy/danish-climat e-and-energy-pol icy/

d a n i s h c l i m a t e c o m m i s s i o n / g r e e n e n e rg y / D o c u m e n t s /green%20energy%20GB%20screen%201page%20v2.pdf (accessed April 30, 2012).

(4) Energi 2050. Udviklingsspor for Energisystemet ; Energinet.dk:Fredericia: Denmark, 2010; http://www.e-pages.dk/energinet/255/fullpdf/full4f900f951a6bd.pdf (accessed April 30, 2012).(5) Tonini, D.; Astrup, T. Life-cycle assessment of biomass-based

energy systems: A case study for Denmark. Appl. Energy 2012 , 99 ,234− 246.(6) Breaking the Biomass Bottleneck of the Fossil Free Society; Concito:

Copenhagen, Denmark, 2010; http://www.fuel.life.ku.dk/Moed-os/∼ /media/Fuel/docs/pdf/Seminar_om_bioenergi_og_baeredygiged/Breaking%20the%20biomass%20bottleneckof%20the%20fossil%20free%20society%20%20%20version%201_220910.ashx (accessed April30, 2012).(7) Mathiesen, B. V.; Lund, H.; Karlsson, K. 100% Renewable energy

systems, climate mitigation and economic growth. Appl. Energy 2011 ,88 , 488− 501.(8) Searchinger, T. D. Biofuels and the need for additional carbon.

Environ. Res. Lett. 2010 , 5 , 024007− 024007.(9) Searchinger, T. D.; Heimlich, R.; Houghton, R. A.; Dong, F.;

Elobeid, A.; Fabiosa, J.; Tokgoz, S.; Hayes, D.; Yu, T. H. Use of U.S.croplands for biofuels increases greenhouse gases through emissionsfrom land-use change. Science 2008 , 319 , 1238− 1240.(10) Indirect Land Use Change from Increased Biofuels Demand.

Comparison of Models and Results for Marginal Biofuels Production from Di ff erent Feedstocks; European Commission Joint Research Centre:Ispra, Italy, 2010; http://ec.europa.eu/energy/renewables/studies/doc/land_use_change/study_4_iluc_modelling_comparison.pdf (ac-cessed April 30, 2012); DOI 10.2788/54137.(11) Bessou, C.; Ferchaud, F.; Gabrielle, B.; Mary, B. Biofuels,

greenhouse gases and climate change. A review. Agron. Sustainable Dev.2011 , 31 , 1− 79.

(12) Dauber, J.; Jones, M. C.; Stout, J. B. The impact of biomass cropcultivation on temperate biodiversity. GCB Bioenergy 2010 , 2 , 289−

309.(13) Valentine, J.; Clifton-Brown, J.; Hastings, A.; Robson, P.;

Allison, G.; Smith, P. Food vs. fuel: The use of land for lignocellulosic’ next generation’ energy crops that minimize competition with primary food production. GCB Bioenergy 2012 , 4 , 1− 19.(14) Hamelin, L.; Joergensen, U.; Petersen, B. M.; Olesen, J. E.;

Wenzel, H. Modelling the carbon and nitrogen balances of direct landuse changes from energy crops in Denmark: A consequential life cycleinventory. GCB Bioenergy 2012 , 4 (6), 889− 907.(15) Environmental Management-Life Cycle Assessment-Principles and

Framework . 2006-07-01, 2nd ed. 2006; ISO 14040.(16) Environmental Management-Life Cycle Assessment-Requirements

and Guidelines. 2006-07-01, 1st ed. 2006; ISO 14040.(17) Spatial Di ff erentiation in Life Cycle Impact Assessment The

EDIP2003 Methodology , Environmental News No. 80; Danish Ministry


dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXXH



of the Environment, Environmental Protection Agency: Copenhagen,Denmark, 2005; http://www2.mst.dk/udgiv/publications/2005/87-7614-579-4/pdf/87-7614-580-8.pdf (accessed April 30, 2012).

(18) Wenzel, H.; Hauschild, M. Z.; Alting, L. Environmental Assessment of ProductsVol. 1: Methodology, Tools and Case Studiesin Product Development ; Kluwer Academic Publishers: Norwell, MA,1997.(19) Jolliet, O.; Margni, M.; Charles, R.; Humbert, S.; Payet, J.;

Rebitzer, G.; Rosenbaum, R. IMPACT 2002+: A new life cycle impactassessment methodology. Int. J. Life Cycle Assess. 2003 , 8 , 324− 330.(20) Pre’ Consultants Simapro 7.2. www.pre.nl (accessed April 30,

2012).(21) Weidema, B.; Frees, N.; Nielsen, A. M. Marginal production

technologies for Life Cycle Inventories. Int. J. Life Cycle Assess. 1999 , 4 ,48− 56.(22) Market Information in Life Cycle Assessment , Environmental

Project No. 863; Danish Ministry of the Environment, EnvironmentalProtection Agency: Copenhagen, Denmark, 2003; www2.mst.dk/udgiv/publications/2003/87-7972-991-6/pdf/87-7972-992-4.pdf (ac-cessed June 1, 2012).(23) Forudsaetninger for Samfundsoekonomiske Analyser Paa Ener-

giomraade; Energistyrelsen, Danish Energy Agency (DEA): Copenha-gen, Denmark, 2010; http://www.ens.dk/da-DK/Info/TalOgKort/Fremskrivninger/beregningsforudsatninger (accessed April 30, 2012).(24) Hamelin, L.; Wesnaes, M.; Wenzel, H.; Petersen, B. M.

Environmental consequences of future biogas technologies based onseparated slurry. Environ. Sci. Technol. 2011 , 45 , 5869− 5877.(25) Fruergaard, T.; Hyks, J.; Astrup, T. Life-cycle assessment of

selected management options for air pollution control residues from waste incineration. Sci. Total Environ. 2010 , 408 , 4672− 4680.

(26) Denmark ’ s National Inventory Report 2011 - Emission Inventories1990− 2009 - Submitted under the United Nations Framework Convention on Climate Change and the Kyoto Protocol; NationalEnvironmental Research Institute: Aarhus University: Aarhus, Den-mark, 2011; http://www2.dmu.dk/pub/fr827.pdf (accessed April 30,2012).(27) Schmidt, J. H. System delimitation in agricultural consequential

LCA - Outline of methodology and illustrative case study of wheat inDenmark. Int. J. Life Cycle Assess. 2008 , 13 , 350− 364.(28) Dalgaard, R.; Schmidt, J. H.; Halberg, N.; Christensen, P.;

Thrane, M.; Pengue, W. A. LCA of soybean meal. Int. J. Life Cycle Assess. 2008 , 13 , 240− 254.

(29) Kloeverpris, J. H. Consequential life cycle inventory modellingof land use induced by crop consumption. PhD thesis: Department of Management Engineering, Technical University of Denmark (DTU),2008, ISBN: 978-87-90855-69-7; http://orbit.dtu.dk/en/publications/consequential-life-cycle-inventory-modelling-of-land-use-induced-by-crop-consumption%286457fc93-c53a-4beb-83ff -015c8047b682%29.html (accessed September 26, 2012).(30) Biofuels in the European Context: Facts and Uncertainties;

European Commission Joint Research Centre: Ispra, Italy, 2008;http://ec.europa.eu/dgs/jrc/downloads/jrc_biofuels_report.pdf (ac-cessed April 30, 2012).(31) Land Use Changes and Consequent CO2 Emissions Due to US Corn

Ethanol Production: A Comprehensive Analysis; Department of Agricultural Economics: Purdue University: Indiana, USA, 2010;https://www.gtap.agecon.purdue.edu/resources/download/5200.pdf (accessed April 30, 2012).(32) Assessing the Land Use Change Consequences of European Biofuel

Policies; International Food Policy Institute (IFPRI): Washington, DC,2011; http://trade.ec.europa.eu/doclib/docs/2011/october/tradoc_ 148289.pdf (accessed April 30, 2012).(33) Plantedirektoratets bekendtgoerelse nr. 786 af 22. juli 2008 om

jordbrugets anvendelse af goedning og om plantedaekke i planteper-iode 2008/2009. Public Law, 2008.https://www.retsinformation.dk/Forms/R0710.aspx?id=116486&exp=1 (accessed April 30, 2012).(34) Bekendtgoerelse om husdyrbrug og dyrehold for mere end 3

dyreenheder, husdyrgoedning, ensilage m.v., BEK nr. 1695 af 19.

december 2006. Public Law, 2006.https://www.retsinformation.dk/Forms/R0710.aspx?id=13063&exp=1 (accessed April 30, 2012).(35) Boerjesson, P.; Berglund, M. Environmental systems analysis of

biogas systems Part I: Fuel-cycle emissions. Biomass Bioenerg 2006 ,30 , 469− 485.(36) Life Cycle Inventories of Bioenergy; Data v 2.0 (2007), Ecoinvent

report No. 17; Swiss Centre for Life Cycle Inventories, ESU-services:Uster, Switzerland, 2007.(37) Emissions from Decentralised CHP Plants 2007 , Environmental

project no. 07/1882. Project report 5− Emission factors and emissioninventory for decentralised CHP production; National EnvironmentalResearch Institute: Aarhus University: Aahrus, Denmark, 2010; http:// www.dmu.dk/Pub/FR786.pdf (accessed April 30, 2012).

(38) A Model for the Carbon Dynamics in Agricultural, Mineral Soils;Faculty of Agricultural Sciences, Aarhus University: Aarhus, Denmark,2010. http://xwww.agrsci.dk/c-tool/NationalModelling.pdf (accessed April 30, 2012).

(39) Petersen, B. M.; Olesen, J. E.; Heidmann, T. A flexible tool forsimulation of soil carbon turnover. Ecol. Model. 2002 , 151 , 1− 14.(40) Cencic, O.; Rechberger, H. Material Flow Analysis with

Software STAN. Environ. Manage. 2008 , 18 (1), 3− 7.(41) Modelling of EnergyCrops in Agricultural Sector Models - A

Review of Existing Methodologies; European Commission Joint ResearchCentre: Ispra, Italy, 2008; http://ftp.jrc.es/EURdoc/JRC42597.pdf (accessed September 26, 2012).(42) GTAP Website. https://www.gtap.agecon.purdue.edu/default.

asp# 2 (accessed September 26, 2012).(43) Mu ̈ller-Wenk, R.; Brandao, M. Climatic impact of land use in

LCA Carbon transfers between vegetation/soil and air. Int. J.LifeCycle Assess. 2010 , 15 , 172− 182.(44) Cherubini, F.; Peters, G. P.; Berntsen, T.; Stromman, A. H.;

Hertwich, E. CO2 emissions from biomass combustion for bioenergy: Atmospheric decay and contribution to global warming. GCB Bioenergy 2011 , 3 , 413− 426.

(45) Drewer, J.; Finch, J. W.; Lloyd, C. R.; Baggs, E. M.; Skiba, U.How do soil emissions of N2O, CH4 and CO2 from perennial

bioenergy crops diff er from arable annual crops? GCB Bioenergy2012 ,4 , 408-419.(46) Joergensen, U. Genotypic variation in dry matter accumulation

and content of N, K and Cl in Miscanthus in Denmark. Biomass Bioenergy 1997 , 12 , 155− 169.

(47) 2006 IPCC Guidelines for National Greenhouse GasInventories. In Agriculture, Forestry and Other Land Use , Chapter 2:Generic Methodologies Applicable to Multiple Land-Use Categories;Intergovernmental Panel on Climate Change: Hayama, Japan, 2006; Vol. 4 , www.ipccnggip.iges.or.jp/public/2006gl/index.html.

(48) Directive 2009/28/EC of the European Parliament and of theCouncil of 23 April 2009 on the promotion of the use of energy fromrenewable sources and amending and subsequently repealingDirectives 2001/77/EC and 2003/30/EC. O fficial journal of the European Union 2009 , 52 , L 140/16− L 140/16-L140/62; http://eur-l e x . e u r o p a . e u / L e x U r i S e r v / L e x U r i S e r v . d o ? u r i =OJ:L:2009:140:0016:0062:en:PDF (accessed April 30, 2012).(49) Plevin, R. J.; O’ Hare, M.; Jones, A. D.; Torn, M. S.; Gibbs, H. K.

Greenhouse gas emissions from biofuels’ indirect land use change areuncertain but may be much greater than previously estimated. Environ.Sci. Technol. 2010 , 44 , 8015− 8021.(50) Nguyen, T. L. T.; Hermansen, J. E.; Mogensen, L. Environ-

mental consequences of different beef production systems in the EU. J.Clean. Prod. 2010 , 18 , 756− 766.(51) Forster, P.; Ramaswamy, V.; Artaxo, P.; Berntsen, T.; Betts, R.;

Fahey, D. W.; Haywood, J.; Lean, J.; Lowe, D. C.; Myhre, G.; Nganga, J.; Prinn, R.; Raga, G.; Schulz, M. Van Dorland, R. Changes inatmospheric constituents and in radiative forcing. In Climate Change2007: The Physical Science Basis. Contribution of Working Group I to theFourth Assessment Report of the Intergovernmental Panel on ClimateChange; Solomon, S., Qin, D., Manning, M., Chen, Z.; Marquis, M., Averyt, K. B., Tignor, M. Miller, H. L., Eds; Cambridge University Press: Cambridge,, 2007; Vol. 51 , pp 130− 234.


dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXXI



(52) O’ Hare, M.; Plevin, R. J.; Martin, J. I.; Jones, A. D.; Kendall, A.;Hopson, E. Proper accounting for time increases crop-based biofuels’

greenhouse gas deficit versus petroleum. Environ. Res. Lett. 2009 , 4 ,

024001.(53) Kendall, A. Time-adjusted global warming potentials for LCA and carbon footprints. Int. J. Life Cycle Assess. 2012 , 17 , 1042− 1049.(54) Styles, D.; Jones, M. B. Energy crops in Ireland: Quantifying the

potential life-cycle greenhouse gas reductions of energy-cropelectricity. Biomass Bioenergy 2007 , 31 , 759− 772.(55) Fazio, S.; Monti, A. Life cycle assessment of different bioenergy

production systems including perennial and annual crops. Biomass Bioenergy 2011 , 35 , 4868− 4878.

(56) Brandao, M.; Mila i Canals, L.; Clift, R. Soil organic carbonchanges in the cultivation of energy crops: Implications for GHG balances and soil quality for use in LCA. Biomass Bioenergy 2010 , 35(6), 2323− 2336.


dx.doi.org/10.1021/es3024435 | Environ. Sci. Technol. XXXX, XXX, XXX− XXXJ



III

Life-cycle assessment of a waste refineryprocess for enzymatic treatment of

municipal solid waste

Tonini, D., Astrup, T.

Waste Management, 32, 165-176



Erratum

Erratum to ‘Life-cycle assessment of a waste renery process for enzymatictreatment of municipal solid waste’ [Waste Management 32/1 (2012), 165–176]

D. Tonini ⇑ , T. AstrupDepartment of Environmental Engineering, Technical University of Denmark, DTU-Building 115, DK-2800 Kongens Lyngby, Denmark

The publisher regrets that Table 7 within the above article con-tained incorrect headings.

The term ‘‘Parameter assessed’’ should have been removed andthe headings of columns 5, 6, 7 and 8 were incorrect and shouldhave read as follows:

Column 5 should be ‘‘PL REC.’’Column 6 should be ‘‘CH 4 ( ; )’’Column 7 should be ‘‘CH 4 (" )’’Column 8 should be ‘‘RR ( ; )’’The publisher would like to apologise for any inconvenience

caused.

0956-053X/$ - see front matter 2011 Elsevier Ltd. All rights reserved.doi: 10.1016/j.wasman.2011.12.009

DOI of original article: 10.1016/j.wasman.2011.07.027⇑ Corresponding author.

E-mail addresses: [email protected] , [email protected] (D. Tonini).

Waste Management 32 (2012) 1054


Waste Management

j ou rna l homepage : www.e l sev ie r. com/ loca te /wasman



Life-cycle assessment of a waste renery process for enzymatic treatmentof municipal solid waste

D. Tonini ⇑ , T. AstrupDepartment of Environmental Engineering, Technical University of Denmark, DTU-Building 115, DK-2800 Kongens Lyngby, Denmark

a r t i c l e i n f o

Article history:Received 28 February 2011Accepted 21 July 2011Available online 21 September 2011

Keywords:Waste reneryLiquefactionLCAEnergy recoveryEnzymes

a b s t r a c t

Decrease of fossil fuel dependence and resource saving has become increasingly important in recentyears. From this perspective, higher recycling rates for valuable materials (e.g. metals) as well as energyrecovery from waste streams could play a signicant role substituting for virgin material production andsaving fossil resources. This is especially important with respect to residual waste (i.e. the remains aftersource-separation and separate collection) which in Denmark is typically incinerated. In this paper, a life-cycle assessment and energy balance of a pilot-scale waste renery for the enzymatic treatment of muni-cipal solid waste (MSW) is presented. The renery produced a liquid (liqueed organic materials andpaper) and a solid fraction (non-degradable materials) from the initial waste. A number of scenariosfor the energy utilization of the two outputs were assessed. Co-combustion in existing power plantsand utilization of the liquid fraction for biogas production were concluded to be the most favourableoptions with respect to their environmental impacts (particularly global warming) and energy perfor-mance. The optimization of the energy and environmental performance of the waste renery was mainlyassociated with the opportunity to decrease energy and enzyme consumption.

2011 Elsevier Ltd. All rights reserved.

1. Introduction

Recovery of material resources and energy from mixed munici-pal solid waste (MSW) has obtained an increased level of attentionin the last decade. While recovery of recyclable materials frommore well-dened waste types, such as: source segregated waste,co-mingled waste or specic industrial/commercial wastes, is pos-sible and typically carried out with good results, material recoveryfrom residual household waste left after source segregation is oftendifcult due to the material properties of the waste (wet, mixedmaterials). Residual household waste may alternatively be inciner-ated and in countries such as Denmark, Sweden, Germany, Austria,and The Netherlands this occurs with very high energy recoveryrates. In Denmark, waste incineration contributes with about 5%of thenational electricity production and 20%of the district heating(Astrup et al., 2009). While residual waste may be a signicantcontributor to energy production, recovery of material resourcesis often limited to the extraction of metals (magnetic and non-magnetic) from the ashes produced. Furthermore, waste incinera-tors must operate as stable as possible throughout the year withlimited options for adjusting energy production to the demandsof society (Fruergaard et al., 2010a ). In the futurewith an increasingshares of uctuating energysources in the energysystem(e.g. wind

power) and greater needs for recovery and recycling of materialsdue to resource scarcity, integrated technologies which both allowmaterial recovery and exible energy production from ‘‘difcult’’waste types such as mixed MSW are considered to be essential.

Waste reneries are one example of a group of integrative tech-nologies which could process mixed residual waste types and pro-duce a range of valuable outputs, including both recyclablematerialsandfuelsforexibleenergyproduction.Onlyafewstudieshave yet provided an evaluation of waste renery processes andthese studies have mainly focused on agricultural waste: in Larsenet al. (2008), a large-scale plant for bioethanol production from lig-nocellulosic biomass was presented. In Lohrasby et al. (2010), a cit-rus waste biorenery was described. Papatheofanous et al. (1995)detailed a pilot-plant biorenery for agricultural residues. Moststudies have focused on lab-scale experiments for the optimizationof the biorenery processes: in Kaparaju et al. (2009), the yields of different products from a rapeseed-based biorenery were investi-gated in a lab-scale setup. Talebnia et al. (2010) presented an over-view of pretreatment, hydrolysis and fermentation experiences forbioethanolproduction fromwheat straw. Examples of outputs fromwaste reneries are solid and liquid fuels as well as recoverablematerials, such as metals, plastic, and nutrients. Waste reneriesmaybesimilarinconcepttobioreneries( Jensen et al., 2010),whichalso produce a range of valuable outputs from a biomass input, andprocess similarities may exist between the two types of reneries.

In order to assess whether waste reneries are environmentallybenecial for treating residual waste compared with existing

0956-053X/$ - see front matter 2011 Elsevier Ltd. All rights reserved.doi:10.1016/j.wasman.2011.07.027

⇑ Corresponding author.E-mail addresses: [email protected], [email protected] (D. Tonini).

Waste Management 32 (2012) 165–176


Waste Managementj ou rna l homepage : www.e l sev ie r. com/ loca te /wasman



treatment options such as incineration with post sorting of metals,a holistic and systematic assessment of both direct and indirectenvironmental impacts is required. Life cycle assessment is a use-ful tool for this. Several LCA studies have focused on bioreneriesfor corn, agricultural waste and other lignocellulosic biomasses(e.g. straw and grass): in the most recent studies, Cherubini andUlgiati (2010) evaluated the environmental impacts using life-cycle assessment (LCA) and other sustainability indicators for abiorenery concept converting wheat straw and corn stover intoethanol, biogas, and other marketable chemical products. Uihleinand Schebek (2009) evaluated the environmental burdens of alignocellulose feedstock biorenery compared to fossil fuels systems.Cherubini and Stroemann (2010) used a Matrix Algebra approachto assess the efciency of rening different types of lignocellulosicbiomass. In Cherubini and Jungmeier (2010), the focus was insteadon a switchgrass biorenery. Techno-economical evaluations werereported (among others) by Mao et al. (2010) and Villegas andGnansounou (2008) . However, the main focus of all these studies

was agricultural residues and other biomasses. No relevant studieson solid waste reneries were found in the scientic literature.The objective of this paper was to evaluate the environmental

sustainability using life-cycle assessment of a specic waste ren-ery concept in which organic waste materials are liqueed usingenzymes and recoverable materials are separated out in a ‘‘solidfraction’’. The assessment is based on experiments at a pilot-scalefacility and includes a range of scenarios with specic focus on en-ergy production and material recycling. This includes evaluation of (i) mass and energy ows in the renery processes, (ii) environ-mental impacts related to the renery processes itself, and (iii)the overall sustainability aspects including downstream use of recovered materials and produced energy.

2. Methodology

2.1. Waste composition

The waste renery process targeted residual municipal solidwaste (MSW), i.e. the fraction remaining after source segregationin the households of the recyclable fractions such as paper, glass,metal, and plastic. As source segregation is not 100% efcient, recy-clable materials were still present in the residual fraction. How-ever, if organic waste is not source segregated the residualfraction proves to be wet and difcult to sort at high efciencies.In Denmark, all residual MSW is collected separately and sent towaste incineration. The waste renery process reported focusedon treating this waste as an alternative to waste incineration,thereby recovering recyclable materials from a waste fraction thatwas not previously subjected to sorting prior to incineration. Thewaste composition has been shown in Table 1 alongside averagevalues for similar waste in Europe.

2.2. Waste renery

The assessment was based on a pilot-scale facility established ata Danish incinerator (Amagerforbrænding, DK). The waste reneryhad a treatment capacity of 1 ton ww/h and had been in operationfrom January 2010 (see Fig. 1). The renery process consisted of two reactors: (1) in the rst reactor the waste was heated by steaminjection to about 95 C for approximately 0.5 h, then cooled toabout 50–55 C before entering the second reactor. (2) In the sec-ond reactor enzymes were added resulting in hydrolysis andbreak-down of bonds in the organic materials thereby essentiallysuspending organic materials in the liquid phase. After the secondreactor, the liquid phase was separated from the remaining solids.

Waste was fed through a hopper to the rst reactor which was acylindrical drum. Water was added to maintain sufcient moisturecontent, and improve mixing of the waste. The heated waste wasdewatered after the rst reactor (water content of about 60–70%by weight), thereby allowing partial recirculation of water in thereactor. The second reactor was also a cylindrical drum. The resi-dence time in this reactor was within 8–16 h depending on processconguration. Twenty-four kilograms of enzymes were added perton of waste (this value can change as a consequence of the processoptimization). A detailed description of the enzymatic processingcan be found in Jensen et al. (2010). Both reactors were insulatedto minimize heat loss and equipped with auxiliary electrical heat-ing for compensation. After the second reactor, separation of liquidand solid fractions was done by a sequence of vibrating screenswhich included: washing of the solid fraction to improve the sep-aration. The energy consumption for operation of the rst reactorwas 1.1 kWhel/ton ww (4 MJ/ton ww), addition of steam (5 bar) ac-counted for another 580 MJ th /ton. Corresponding energy of 4.7 kWhel/ton ww (17 MJ/ton ww) was consumed for the dewater-ing of the waste by a vibrating screen and water recirculation. Sub-

sequent cooling of the waste accounted for about 4.7 kWh el/ton ww. The energy consumption for operation of the second reac-tor was 5.3 kWhel/ton ww (19 MJ/ton ww) while the following sep-aration of the output into a liquid and solid fraction amounted to afurther 4.7 kWh el/ton ww. The nal sieving included the pressingof the solid material to increase the recovery rate of the liquid frac-tion as well as dewatering the solids. All the energy data was basedon process data from the operation of the pilot-scale plant onresidual municipal solid waste. In total, the waste renery processconsumed about 33 kWh el/ton ww and 580 MJth /ton ww (as heat).

The outputs from the waste renery process per 1 ton of treatedresidual waste were 1023 kg of liquid fraction (total solids (TS)30%) and 470 kg of solid fraction (57% TS). The chemical composi-tion of liquid and solid fraction is reported in Table 2. These valuesare to be considered as a result of preliminary investigations andcan change following the future plant development and process

optimization. The liquid fraction consisted primarily of suspendedorganic matter (food waste and paper), while the solid fractionmainly consisted of non-degradable materialssuch as glass, plastic,metals, textiles, soil, and ceramics. In these pilot-scale experi-ments, recyclable materials were hand-sorted from the solid frac-tion with the following estimated recoveries: glass (85%), ferrousmetals (85%), and non-ferrous metals (60%, assumed to be alumin-ium). Similar recoveries have been reported in other studies onwaste (e.g. Rigamonti and Grosso, 2009; Arena et al., 2003). Energyconsumption for the separation of recyclable materials was esti-mated at 18 kWh el/ton waste treated. Upgrading of the liquid frac-tion for fuel quality prior to co-combustion in power plants (seelater description of assessment scenarios) was carried out by dry-ing the materials reaching a caloric value of 14.5–15 GJ/ton (85%TS). The drying was estimated to consume additional 402 MJth /ton ww.

Table 1

Waste composition used in this study (% of each fraction per wet weight).

Fraction Average Danishresidual waste

Average European MSW

Riber et al. (2009) Christensen et al. (2009)

Organic 45.1 35Paper 12.1 22Plastic 9.2 10Glass 2.9 6Metal 3.5 4Textile 1.9 3Other 25.3 20

Sum 100 100

166 D. Tonini, T. Astrup / Waste Management 32 (2012) 165–176



2.3. Goal and scope of the assessment

The functional unit of the life cycle assessment (LCA) was

‘‘treatment of one ton (1000 kg) of Danish residual municipal solid(wet) waste’’. The chemical composition of the waste was assumedto be identical to the average Danish residual household waste asreported by Riber et al. (2009).

The assessment considered the waste as a ‘‘zero-burden’’boundary (i.e. the waste as such was not assumed to carry anyenvironmental impacts). Downstream utilization of recoveredheat/electricity and recyclable materials were credited the systemby system expansion into the energy and industrial sectors (savedproduction of energy and virgin materials). The boundary of thesystem was set at the treatment facility gate (incinerator andwaste renery) and included downstream disposal of incinerationashes and recycling of materials. The impacts related to waste col-lection were excluded from the assessment as they contributedequally to all scenarios. Also the environmental impacts associatedwith the construction and demolitions of facilities were not in-

cluded. Following common practices in LCA studies, all environ-mental impacts (resource consumption, emissions to air, soil andwater) related to the transportation and treatment (until nal dis-posal) of the liquid and solid fractions, recyclables, bottom, and yashes were included for a time horizon of 100 years. An overviewof the LCA system boundary is presented in Fig. 2.

2.4. Impact assessment

The assessment was carried out according to the LCA methodEDIP 1997 (Wenzel et al., 1997) following the principles of conse-quential LCA (e.g. Finnveden et al., 2009), i.e. focusing on the conse-quences of a decision(in this case treatmentof thewastepreviouslymentioned). This for example means that energy generated by thewaste system was assumed to substitute energy production at theplants which actually respond to the change, rather than substitut-

ing average energy production ( Weidema et al., 1999). When sys-tem expansion was not applicable, for instance in the case of cogeneration of heat and electricity at combined heat and powerplants (CHP), allocation based on energy quality (exergy) was in-stead applied according to that of Fruergaard et al. (2010a).

The following impact categories were included in the assess-ment: Global Warming (GW), Acidication (AC), Nutrient Enrich-ment (NE), Ecotoxicity in water chronic (ETwc), Human Toxicityvia water (HTw), Human Toxicity via soil (HTs), Human Toxicityvia air (HTa).

2.5. LCA scenarios

The life cycle assessment included two sets of scenarios: (1)various approaches for energy utilization of outputs from thewaste renery, and (2) various congurations of the surroundingenergy system reecting a range of assumptions regarding down-stream energy substitution.

2.5.1. Waste renery scenarios

Five different scenarios for the treatment of the residual muni-cipal solid waste were evaluated:

(1) INC: all waste was incinerated. Recycling of ferrous metalsand aluminium separated from incineration ashes.

(2) CC-CC: co-combustion of the liquid and solid fractions afterdrying, andrecycling of glass,ferrous metals, andaluminium.

(3) CC-INC: co-combustion of the liquid fraction after drying,incineration of the solid fraction, and recycling of glass, fer-rous metals, and aluminium.

(4) BG-CC: anaerobic digestion of the liquid fraction, co-com-bustion of the solid fraction after drying, and recycling of glass, ferrous metals, and aluminium. Biogas was assumedcombusted in a gas-red combined heat and power plant,generating heat, and electricity.

(5) BG-INC: anaerobic digestion of the liquid fraction, incinera-

tion of the solid fraction, and recycling of glass, ferrous met-als, and aluminium. Biogas was assumed to be combusted ina gas-red combined heat and power plant, generating heat,and electricity.

2.5.2. Energy system scenariosSpecial attention was devoted to assumptions regarding the

surrounding energy system as choices here may signicantly affectthe outcome of the LCA (Fruergaard et al., 2009; Fruergaard andAstrup, 2011; Finnveden et al., 2009; Ekvall and Weidema, 2004;Weidema et al., 1999). In a short term perspective (e.g. withinthe coming 5–15 years), it can be assumed that the existing energyproduction capacities respond to changes in the waste sectoraccording to their relative share of the energy production. In alonger term perspective (e.g. beyond 15 years), it may be assumedthat energy from waste contributes to the decommissioning of

Fig. 1. The waste renery process (mass and energy ows).

Table 2

Average chemical composition of the liquid and solid fraction (mg/kg DM) obtainedfrom the waste renery. The latter was based on modelling and did not include glassand metals which were assumed to be separated for recycling.

Element Liquid Solida Element Liquid Solida

C 411 503 P 2.2 2.5H 55.8 68.6 As 0 0.01N 16.9 10.1 Ba 0.09 0S 2 0.3 Cd 0 0.004Cl 12 16.9 Cr 0.02 0.2Al 6.9 2.4 Cu 0.02 1.1Ca 25.6 16.3 Hg 2E-4 0.001Fe 2.9 12.1 Mn 0.05 0.04K 7.4 4.0 Ni 0.01 0.04Mg 2 1.3 Pb 0.02 0.3Na 16 6.6 Zn 0.14 1.3

a After separation of recyclables (metals and glass).

D. Tonini, T. Astrup/ Waste Management 32 (2012) 165–176 167



fossil based energy production capacities (both electricity andheat) as these technologies are generally intended to be phased

out in order to comply with political CO2 reduction targets. Of thefossil fuels, coal, and natural gas represent the two ends of the range with respect to CO2 emissions per combustion unit of fuel energy (95 kg CO2/GJ coal and 56.77 kg CO2/GJ natural gas).While electricity production from waste can be consideredmarginal compared with the total electricity production in mostcountries, substitution of district heating often depends on localconditions and production capacities connected to the districtheating network in question ( Fruergaard et al., 2010a). This meansthat when evaluating a system in a short term perspective involv-ing existing production capacities, substitution of district heatingshould reect local conditions. However, it is viable to assume thatin the long term heat production from waste will contribute tophasing-out fossil fuels.

In this study, ve different energy systems were evaluated foreach waste renery scenario. One energy system reected the

current situation (short term) in Copenhagen as an example of a complex district heating system, while the four additional

systems reected a long term perspective with and without heatrecovery:

(1) ST-CPH (Short Term-Copenhagen): electricity from wasteincineration was assumed to substitute electricity produc-tion at coal-red power plants and heat production wasassumed to substitute a mix of the following fuels represent-ing district heating production in the Copenhagen area: coal(11.4%), fuel oil (4.8%), natural gas (18.6%), straw and woodpellets (23.1%). Co-combustion of the liquid and solid frac-tions was in this case assumed to directly replace coal atthe existing CHP plants. Similarly, biogas was assumed todirectly replace natural gas also at CHP plants. Consequently,no changes in energy production from these plants wereassumed and energy substitution only concerned fuel con-sumption at the plants.

Fig. 2. Overview of the system boundary of the LCA study.




(2) LT-CO (Long Term-Coal): electricity and heat productionfrom waste was assumed to substitute production at coal-red CHP plants (allocation according to energy quality).

(3) LT-NG (Long Term-Natural Gas): electricity and heat produc-tion from waste was assumed to substitute energy produc-tion at natural gas red CHP plants (allocation according toenergy quality).

(4) LT-CONH (Long Term-Coal, No Heat): electricity productionfrom waste was assumed to substitute production at coal-red CHP plants (allocation according to energy quality).No heat recovery was assumed.

(5) LT-NGNH (Long Term-Natural Gas, No Heat): electricity pro-duction from waste was assumed to substitute production atnatural gas red CHP plants (allocation according to energyquality). No heat recovery was assumed.

The last two energy scenarios without heat recovery were in-cluded to better illustrate the importance of district heating and

provide scenarios better reecting the situation in countries withsignicantly less heat recovery from waste.

2.6. Associated technologies

2.6.1. Waste incinerationWaste incineration was modelled as a grate-red incinerator

equipped with wet ue gas cleaning, selective non-catalytic reduc-tion (SNCR) of NO x, Hg and dioxin removal by activated carbon.Gross electricity and heat efciencies of the incinerator were 20%and 65%, respectively, relative to the lower heating value (LHV)of the waste input. These are considered average values for Den-mark (Fruergaard and Astrup, 2011 ). Internal electricity consump-tion at the plant was 65 kWh/ton of waste plus an additional0.42 L/ton of oil as auxiliary fuel, and 0.66 kg NaOH/ton and7.85 kg CaCO3/ton for ue gas cleaning (Astrup et al., 2009;Fruergaard and Astrup, 2011 ).

Following the approach of Riber et al. (2009), emissions weredivided into either process-specic emissions (emissions indepen-dent of waste composition but proportional to the amountof wasteincinerated) or waste-specic emissions (determined by outputtransfer coefcients). Selected air emissions are shown in Table 3.One hundred and nineteen kilograms of bottom ashes, 17 kg of yashes and other 12 kg APC (Air Pollution Control) residues weregenerated per ton of waste. Magnetic and non-magnetic metalswere recovered from bottom ashes (overall recovery estimated at38%) prior to utilization as construction material substituting nat-ural gravel (Birgisdottir et al., 2007). APC residues were assumed tobe utilized in the backlling of old mines ( Fruergaard et al., 2010b).

2.6.2. Coal-red CHP plant

Co-combustion of liquid and solid fractions from the wasterenery were modelled as a coal-red CHP plant. The plant wassuspension red with pulverized coal and equipped with semi-dry ue gas cleaning and selective catalytic reduction (SCR) of NO x (Fruergaard and Astrup, 2011 ). Net electricity and heat ef-ciencies of the incinerator were both 40% (DONG, 2008), relativeto the lower heating value (LHV) of the waste input. Reagentsfor ue gas cleaning amounted to 19 kg CaO/ton, 1 kg NH3/ton,0.21 kg NaOH/ton and 0.14 kg HCl/ton (Astrup et al., 2009;Fruergaard and Astrup, 2011 ). Selected air emissions are shownin Table 3. One hundred and seventy ve kilograms of bottomashes, 20 kg of y ashes, and 13 kg of air-pollution-control (APC)residues were generated per ton of waste co-combusted. Fly ashwas assumed to be used for backlling of old salt mines whilethe remaining solid residues were assumed to be landlledfollowing the approach of Fruergaard et al. (2010b).

Prior to co-combustion, the liquid fraction was assumed to bedried to reach a LHV of about 15 GJ/ton with a total solid contentof 85% TS. The energy consumption associated with drying wasestimated to be 112 kWh th /ton ww (402 MJth /ton ww).

2.6.3. Anaerobic digestionThe liquid fraction was assumed to be digested using a one

stage mesophilic anaerobic digestion plant appropriate for organicmunicipal solid waste ( Boldrin et al., 2011). A methane yield corre-sponding to 75% of the methane potential of the liquid fraction wasassumed, in agreement with similar practice for organic waste(Davidsson et al., 2007; Pognani et al., 2009 and Møller et al.,2010). The methane content in the biogas was assumed to be63% (vol.). Internal energy consumption at the plant was: 0.9 L of diesel and 18 kWh of electricity per ton of wet waste received atthe plant. The electricity consumption was considerably lowerthan typical plants treating organic waste because the slurry didnot require typical pre-treatment such as shredding, sieving, plas-

tic/metal removal and hydrolysis. The dry matter content of thedigestate was set to 10% based on a mass balance on the system.Overall, 84 Nm3 CH4 (133 Nm3 biogas) and 2.2 tons of digestate(10% TS) were generated per ton of liquid fraction. Any rejectappearing was neglected. Fugitive methane emissions were as-sumed controlled by appropriate air controls and a are ( Boldrinet al., 2011).

2.6.4. Natural gas-red CHP plant Combustion of biogas generated by anaerobic digestion was as-

sumed to occur in a gas-red combined heat and power plant. Theinventory of resource consumption was assumed to be the same asdescribed for the coal-red power plant. The energy efciency wasassumed equal to the coal-red CHP plant (i.e. 40% electricity and40% heat recovery based on the LHV of the biogas). The main airemissions associated with biogas combustion were: SO 2 = 0.11 g/

Nm3

biogas, NO x = 3 g/Nm3

biogas, CH4 = 1.8 g/Nm3

biogas,N2O = 0.0028 g/Nm3 biogas (Nielsen and Illerup, 2006).

2.6.5. Recycling of glass, metals, and plastic Glass recycling was assumed to substitute 99% virgin produc-

tion through re-melting of cullet and with a market substitutionratio of 100% for the produced glass. Glass recycling included theproduction of glass from cullet minus the avoided virgin produc-tion. Savings in providing virgin resources for glass productionwas not included ( Larsen et al., 2010; DTU Environment, 2008).The benet of glass recycling was primarily related to savings inenergy consumption, corresponding to a new saving of 230 kg CO2-eq/ton of glass input.

Ferrous metal recycling was assumed to substitute 100% virginproduction with a market substitution ratio of 100% for the pro-duced metal. Metal recycling included re-melting of scraps and

rolling of new steel sheets from the melted metal waste, minusthe avoided virgin production ( Larsen et al., 2010; DTU Environ-ment, 2008 ). The benet of metal recycling was primarily savingsin energy consumption, corresponding to a net saving of 1689 kg CO2-eq/ton of metal input.

Aluminium recycling was assumed to substitute 100% virginproduction with a market substitution ratio of 100% for the pro-duced aluminium. An overall material loss of 21.2% was assumeddue to the sorting process. Aluminium recycling included re-melt-ing of aluminium scrap and alloying, minus the avoided virgin pro-duction ( Larsenet al., 2010; DTUEnvironment, 2008). The benetof aluminium recycling was primarily savings in energyconsumption,corresponding to a net saving of 7698 kg CO 2-eq/ton of metal input.

Plastic recycling was assumed to substitute 90% virgin produc-tion by re-melting with a market substitution ratio of 90%. The lat-ter value was a rough estimate of the potential decrease in material




quality. According to Schmidt and Stromberg (2006) , the loss of material quality can be as high as 20%. However, they also stated

that the loss highly depends on the eld of application of the sec-ondary plastic. For instance, in the case that the recovered plastic isutilized as an admixture in the production of primary plastic theremay be no loss. Thus, in this study, 10% loss in material quality wasassumed. Plastic recycling included the granulation and re-meltingfor production of PE plastic from waste plastic minus the avoidedvirgin production ( DTU Environment, 2008). The benet of plasticrecycling was primarily savings in energy consumption, corre-sponding to a net saving of 644 kg CO2-eq/ton of plastic input.

2.6.6. TransportationThe fuel consumption for transport of the liquid and solid frac-

tion as well as of the residues from incineration, combustion andanaerobic digestion (i.e. bottom and y ash, APC, and digestate)was represented by ‘‘transportation’’ to the point of unloading,e.g. at the treatment or disposal facilities. The fuel consumptionwas expressed in fuel consumption per ton of waste per km(one-way-distance) according to Eisted et al. (2009). Transport dis-tances for the liquid and solid fractions between the waste reneryand downstream utilization were assumed to be 15 km based onDanish conditions. Transport distances related to solid residuesfrom incineration and co-combustion were assumed to be 70 kmfor bottom ashes and 500 km for APC residues and y ashes. Asrecyclables generally enter a global market, transport distanceswere unknown in this case but average European values were usedfor approximation ( Eisted et al., 2009): 100 km for glass and plasticand 500 km for aluminium and ferrous metals.


The results of the Life-cycle impact assessment (LCIA) are re-

ported in Figs. 3 and 4 and Tables 5–7. The results of the LCIAare expressed as normalized impact potentials in the unit milliPer-son Equivalent (mPE) per ton ww. EU 15 normalization referenceshave been used in the normalization step ( Table 4). One PE corre-sponds to the environmental load caused by one average EU 15 cit-izen in one year (reference year: 1990) covering all activities in life(mining, agriculture, transport, housing, etc.).

3.1. Environmental performance of the waste renery

Evaluating the waste renery process, without considering thesystem context, can provide information about which sub-pro-cesses are most important. Fig. 3 shows the potential impacts re-lated to Global Warming (GW), Acidication (AC), and NutrientEnrichment (NE) for two of the energy system scenarios: LT-COand LT-NG. Only the environmental categories GW, AC, and NE

were considered for discussion as the other environmental andtoxiccategories proved negligible. Only the two energy system sce-narios LT-CO and LT-NG were addressed. The results can be consid-ered representative also for the other energy scenarios. The mainimpacts associated with GW were related to the production of en-zymes (4.4 kg CO2-eq/kg of enzyme, corresponding to 12 mPE/ton ww treated). Electricity and heat consumption (respectively,equal to 33 kWhel/ton ww and 580 MJth /ton ww) contributed withsmaller loads (the magnitude depended on whether coal or naturalgas was substituted). For the scenarios including co-combustion(CC-CC and CC-INC), drying of the liquid fraction contributed withan additional 11 and 1.5 mPE/ton ww treated in the coal and nat-ural gas scenarios, respectively. This means that the load due to en-ergy consumption in the drying process may be as high as theimpact caused by enzyme production in case of coal energy substi-

tution (i.e. LT-CO).The impacts on the category AC (6–8 mPE/ton ww) were againrelated to the energy consuming processes such as drying of theliquid fraction and industrial production of enzymes (release of SO2 and NO x from combustion).

Table 3

Transfer coefcients to air (% of input transferred to the air emissions) of selectedelements.

Element Transfer coefcientincineration (% TS)

Transfer coefcientco-combustion (% TS)

Al – 0.1As 0.204 0.39Cd 0.006 0.99Cr 0.068 0.17Cu 0.009 0.17Fe – –Hg 3.5 28.5Mg – –Mn 0.004 0.08Ni 0.125 0.29Pb 0.015 0.35Zn – –

Table 4

Environmental impact categories and normalization references included in theassessment ( Stranddorf et al., 2005; Hansen et al., 2004 ).

Acronym Physicalbasis

Normalizationreference EU-15

Unit

Non-toxicity categoriesGlobal warming GW Global 8,700 kg CO2-eq./

person/yrAcidication AC Regional 74 kg SO2-eq./

person/yrNutrient

enrichmentNE Regional 119 kg NO3 -eq./

person/yr

Toxicity categoriesEcotoxicity in

water chronicETwc Regional 352,000 m3 water/

person/yrHuman toxicity

via soilHTs Regional 127 m3 soil/

person/yrHuman toxicity

via waterHTw Regional 50,000 m3 water/

person/yr

m P E / t o n n e w w

Enzymes

Drying

Electricity

Heat

0

5

10

15

20

25

30

LT-CO LT-NG LT-CO LT-NG LT-CO LT-NG

GW AC NE

Fig. 3. Impact potentials (mPE/ton ww) associated with the treatment of oneton ww in the waste renery and (eventual) subsequent drying process (onlyrelevant environmental categories).




Industrial production of enzymes contributed with a signicantenvironmental load on the Nutrient Enrichment (NE) category dueto the release of nutrients in the forms of NO x and phosphates to

water (approximately 15 mPE/ton ww). This was in accordancewith other similar results found in the literature ( Nielsen et al.,2007).

mPE/tonne ww

GW

mPE/tonne ww

ETwc

AC HTw

NE HTs

-120-80-40040 -300-200-100001 0200300

mPE/tonne ww mPE/tonne ww-120-80-40040 -300-200-100001 0200300

INC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INC

BG-CCBG-INC

S S T - C P

H

L T - C

O

L T - N

G

L T - C

O N H

L T - N

G N H

INC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INC

BG-CCBG-INC

S S T - C P

H

L T - C

O

L T - N

G

L T - C

O N H

L T - N

G N H

INCCC-CC

CC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INC

S S T - C

P H

L T - C

O

L T - N

G

L T - C

O N H

L T - N

G N H

INCCC-CC

CC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INC

S S T - C

P H

L T - C

O

L T - N

G

L T - C

O N H

L T - N

G N H

INCCC-CC

CC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INC

S S T - C

P H

L T - C

O

L T - N

G

L T - C

O N H

L T - N

G N H

INCCC-CC

CC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INCINC

CC-CCCC-INCBG-CC

BG-INC

S S T - C

P H

L T - C

O

L T - N

G

L T - C

O N H

L T - N

G N H

Transportation

Waste refinery

Recycling

Energy recovery

Drying process

Use-on-land

Landfill

Fig. 4. Environmental impacts on relevant non-toxic and toxic impact categories: process contributions.




Overall, to optimize the waste renery process focus should be

directed to the consumption of energy and enzymes. 3.2. Short term perspective

The short term perspective reects the current boundary condi-tions in the Copenhagen area and was evaluated as an example of implementing the waste renery concept in an existing districtheating system with signicant shares of biomass. Results for theve waste management scenarios are shown in Fig. 4 (non-toxicand toxic categories). In order to simplify the discussion, the netnumerical values are presented in Table 5.

3.2.1. Non-toxicity impact categoriesImpacts related to GW were in the ve scenarios strongly corre-

lated with the energy recovery at the incinerator or CHP plant andthe corresponding energy substituted (and savings in CO 2 emis-

sions). The co-combustion scenario (CC-CC) appeared to have thebest performance ( 81 mPE/ton ww) due to the high electricityefciency of the CHP plant. As part of the district heating generatedin the incineration scenario (INC) substituted biomass resources,which were considered CO2 neutral within the short term perspec-tive, this scenario appeared less competitive ( 13 mPE/ton ww)compared with the other scenarios. A consequence of substitutingheat production at back-pressure CHP plants (the biomass CHPplants in the Copenhagen area) is a decreased electricity produc-tion at the same plants. This requires that other plants ‘‘deliver’’the electricity decit to maintain system comparability therebyinducing an additional environmental load ( Fruergaard et al.,2010a). It was assumed that this electricity was marginal and pro-duced based on coal, following the approach of Fruergaard et al.(2010a). An important assumption for the context was thatbiomass resources were not considered constrained. Thus, the

biomass resources saved by substitution of heat at biomass plants

were not assumed to be used elsewhere to off-set fossil fuels. Thiscan be debated; however, the assumption was used here to illus-trate a situation in which biomass was readily available on aninternational market. The alternative situation in which biomassis a constrained resource, and any biomass saved will thus be usedelsewhere to off-set fossil fuels, is covered by the long term per-spective in this paper (the fossil fuels being either coal or naturalgas). The difference between energy recovery for the INC scenarioin the short term perspective and the long term perspective there-by illustrates the importance of this assumption.

The importance of co-combusting the solid fraction relative toincineration was illustrated by the difference found between theCC-CC and CC-INC scenarios, i.e. about 41 mPE/ton ww. The biogasscenario (BG-CC) with co-combustion of the solid fraction showedsavings similar to the CC-INC scenario ( 36 mPE/ton ww). As withthe scenario in which the liquid fraction is co-combusted, it was

found that again incinerating the solid fraction resulted in smallersavings (BG-INC) compared to co-combustion.The results for AC followed the same overall trend as GW be-

cause the relevant emissions (nitrogen oxides and sulphur dioxide)are also related to energy production. Nutrient Enrichment (NE)showed a different trend as application of digestate on soil afteranaerobic digestion of the liquid fraction led to signicant loads.This is because use of digestate on land increased NO3 andPO4

3 leaching compared with application of inorganic fertilizers.This was in agreement with the results of other studies on anaer-obic digestion (Sander et al., 2003).

3.2.2. Toxicity impact categoriesThe impacts related to Ecotoxicity in water (ETwc) were mainly

related to the recycling rates of aluminium (60%) and ferrous met-als (85%). The higher recycling rates found in the waste renery

Table 5

Environmental impacts on non-toxic and toxic categories: overall results (mPE/ton ww).

Category Scenario Energy scenario

ST-CPH LT-CO LT-NG LT-CONH LT-NGNH

GW INC 13 67 16 29 7CC-CC 81 106 44 80 38CC-INC 40 81 28 68 20BG-CC 36 64 16 52 16BG-INC 4 40 1 16 1

AC INC 5 13 1 3 2CC-CC 12 22 7 16 6CC-INC 9 19 5 15 4BG-CC 8 16 5 13 5BG-INC 5 13 3 8 3

NE INC 7 3 8 7 10CC-CC 9 1 10 7 11CC-INC 14 5 14 8 15BG-CC 49 43 50 46 50BG-INC 53 47 54 52 54

ETwc INC 235 157 143 152 143CC-CC 189 210 187 207 187CC-INC 255 227 208 225 208BG-CC 184 199 183 197 183BG-INC 251 217 206 214 206

HTw INC 107 96 108 101 109CC-CC 175 174 192 177 192CC-INC 19 13 28 15 28BG-CC 193 187 200 189 200BG-INC 37 26 35 29 35

HTs INC 29 26 22 24 21CC-CC 17 18 11 17 11CC-INC 74 73 68 72 68BG-CC 212 210 214 210 214BG-INC 155 155 158 155 158




scenarios induced greater savings compared with incineration(INC). This was principally due to avoided emissions from virginmetal production (Polycyclic Aromatic Hydrocarbons, i.e. PAHs,Fe, Cd, Sr) as well as the associated energy savings (especially withrespect to aluminium production).

For Human Toxicity via soil (HTs), the environmental impactsprimarily originated from As and Hg emissions to soil from theapplication of digestate to farmland (scenarios BG-CC and BG-INC). Also, to a smaller extent from Hg emissions to air from thecombustion processes. Hg emissions to air were also the main con-tribution responsible for impacts related to Human Toxicity viawater (HTw). In this category, better ue gas cleaning at the wasteincinerator (including Hg removal) allowed an improved environ-mental performance compared with the scenarios involving co-combustion of the solid fraction (CC-INC, BG-INC).

3.3. Long term perspective

The long term perspective is not related to a specic districtheating network or a specic geographic area. Instead, it reectsa potential future situation in which waste contributes to the tran-sition away from fossil fuels by substituting fossil fuels such as coalor natural gas. The results are shown in Fig. 4. In Table 5 the netnumerical results are reported.

3.3.1. Non-toxicity impact categoriesThe performance of the ve assessed scenarios with respect to

GW was again related to energy recovery (particularly electricity)at the power plant orthe incinerator,as shown in Fig. 4and Table 5.The assumption regarding the marginal energy substituted (eithercoal or natural gas) played a critical role with respect to the mag-nitude of the results. However, this choice did not change the over-all ranking of the scenarios. The scenario in which the solid and

liquid fraction was co-combusted (CC-CC) achieved the best envi-ronmental performance (savings equalled 106 and 44 mPE/ton ww in LT-CO and LT-NG, respectively). The scenario which in-cludes anaerobic digestion of the liquid fraction and co-combus-tion of the solid fraction (BG-CC) gave comparable results withthe reference scenario (incineration). In fact, even though the en-ergy efciency was found to be higher in the scenario involvingbiogas production and co-combustion (BG-CC) as demonstratedin the energy balance in Table 6, the environmental load causedby the waste renery itself decreased the overall benets. Thiswas found to be the case for all the options including the reningprocess. Finally, the lower electricity recovery at the incineratormade the option of incinerating the solid fraction less attractivecompared to co-combustion in power plants ( Fig. 4).

As already discussed for the short-term perspective, the trendfor AC followed the results reported for the GW category. However,the NE impact category proved to be contrary to this as the appli-cation of digestate on land led to signicant impacts due to in-creased leaching of nutrients compared to the use of inorganicfertilizers.

3.3.2. Toxicity impact categoriesThe benets found with the ETwc impact category ( Fig. 4) were

principally related to the recycling rates of aluminium and ferrousmetals. As already discussed for the short-term perspective, thehigher efciency of recycling in the waste renery induced higherenvironmental savings compared with incineration. This was thecase for all options including the rening process.

Signicant environmental loads (approximately 155–214 mPE/ton ww), associated with the scenarios in which the liquid fractionis sent to anaerobic digestion (BG-CC and BG-INC), were found forthe Human Toxicity via soil (HTs) category. The impacts were pri-marily caused by As and Hg emissions to soil from digestate appli-cation. However, this data was determined by modelling inEASEWASTE and experimental data is needed in order to evaluatethe quality of the digestate in more detail.

Finally, with regards to the HTw category, it was found that bet-ter removal of Hg during the ue gas cleaning in the incinerator al-lowed for an improved environmental performance for theapplicable scenarios, for example, the incineration of the solid frac-tion (CC-INC and BG-INC) compared to those scenarios in whichthe solid fraction is co-combusted (CC-CC and BG-CC).

3.4. The signicance of heat production

The co-generation of heat and electricity is strictly connected tolocal conditions such as the presence of a district heating network

and an existing demand for heat throughout the year. Conse-quently, two management scenarios without heat production(LT-CONH and LT-NGNH) were evaluated in order to show the sig-nicance of heat production on the overall results. In general theresults showed that, whenever heat was not co-generated or uti-lized in the incinerator or power plant, all waste renery manage-ment scenarios achieved a better environmental performancecompared to incineration (INC), see Fig. 4 and Table 5. This wasfound to be especially true with respect to GW ( Fig. 4). The oppor-tunity of generating a high quality energy carrier, such as electric-ity, from solid and liquid fractions, was associated with signicantenvironmental savings in all the scenarios including the wasterenery (CC-CC and BG-CC in particular). The magnitude of theenvironmental benets was higher when substituting coal-based

Table 6

Energy balance of the ve scenarios. The results are reported as GJ of primary energy/ton ww (rounded values) for both coal- and natural gas-based energy systems. Energyconsumption is expressed as positive value while avoided energy consumption as negative.

Process INC CC-CC CC-INC BG-CC BG-INC

Coal NG Coal NG Coal NG Coal NG Coal NG

Transportation 0.3 0.3 0.3 0.3 0.3 0.3 0.2 0.2 0.2 0.2Waste renery – – 3.8 3.4 3.8 3.4 3.8 3.4 3.8 3.4Drying process – – 1.9 0.3 1.9 0.3 – – – –INC – El substitution 7.7 7 – – 4.0 3.6 – – 4.0 3.6INC – heat substitution 4.2 0.8 – – 2.2 0.5 – – 2.2 0.5CHP – El substitution – – 17.7 16 9.3 8.4 11.7 9.1 5.3 1.5CHP – Heat substitution – – 3.1 0.8 1.6 0.4 2.0 0.6 0.5 0.3Ash treatment 0.04 0.03 0.02 0.02 0.02 0.01 0.01 0.01 0.01 0.01Al Recycling 0.3 0.3 0.7 0.7 0.7 0.7 0.7 0.7 0.7 0.7Glass Recycling 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1Fe Recycling 0.2 0.2 0.3 0.3 0.4 0.4 0.3 0.3 0.4 0.4Substitution of inorganic fertilizers – – – – – – 0.2 0.2 0.2 0.2

Total 12.1 8.0 15.9 13.9 12.3 10.1 11.6 7.5 9.4 3.7




energy than natural gas-based, a consequence of the CO 2 emissionsassociated with the production and combustion of the fuel.With regards to AC, the trend was the same as for GW. However,

this pattern was not found with the other environmental impactcategories. The other results including the ranking of the scenarioswere similar to the energy scenarios assuming heat production.

3.5. Energy balance

An energy balance was carried out in order to evaluate the en-ergy efciency and related performance of the assessed wastemanagement scenarios. The balance accounted for all the energy-related inputs and outputs (electricity, heat, fuels, including energyrequired to extract and produce the fuel) to and from the systemsduring the entire life-cycle of the waste. The heat and electricitygenerated from the scenarios were assumed to substitute coal or

natural gas marginal heat/electricity. In case of CHP plants, the en-ergy recovered in the form of heat was calculated based on theexergy content (0.15), in accordance with Fruergaard et al.(2009). Substitution of fertilizers and related energy savings werealso included.

As shown in Table 6, the ranking of the scenarios followed theresults achieved for the Global Warming category (GW), conrm-ing that energy recovery (particularly electricity) was most impor-tant to GW, as previously discussed. The best energy performance( 15.9 GJ/ton ww in the coal scenario and 13.9 GJ/ton ww in thenatural gas scenario) was achieved with co-combustion of the twofractions (CC-CC). The scenario including anaerobic digestion of theliquid fraction and co-combustion of the solid fraction (BG-CC)achieved similar performances of incineration ( 11.6 GJ/ton wwin the coal scenario and 7.5 GJ/ton ww in the natural gasscenario).

As shown in Table 6, energy recovery at the CHP plant was themost important energy saving whereas the most important energyexpenses were related to the waste renery and to the (eventual)drying process. The waste renery contributed with an energy con-sumption of between 3.4 and 3.8 GJ/ton ww depending on the fuelsubstituted. The production of enzymes was the most energy-intensive process (about 1.4 GJ/ton ww). The drying process wasestimated to require between 0.3 and 1.9 GJ/ton ww dependingon fuel substituted. Among the recycling processes, aluminiumrecycling contributed with the most signicant energy savings(about 0.7 GJ/ton ww).

3.6. Sensitivity analysis

A sensitivity analysis was produced in order to evaluate theinuence of relevant assumptions and parameters in the assess-

ment. The analysis focused on the following key assumptions of the study: waste composition, impacts related to industrial pro-duction of enzymes, potential for plastic recycling, methane poten-tial of the liquid fraction, recycling rates in the waste renerysystem and utilization of the biogas in more efcient integratedgas combined cycle (IGCC) facilities. For the sensitivity analysis,only the results for the energy system LT-CO based on coal as mar-ginal energy were reported ( Table 7). This was because the resultsindicated the same trend for all the other energy scenarios.

A different waste composition ( Christensen et al., 2009) wasused in order to assess the performance of the scenarios with awaste composition typical of MSW without source-segregation(Table 1). Consequently, this allowed an assessment of the inu-ence of a higher content of metals and paper in the total mixedwaste on theenvironmental impactcategories. Theresults ( Table 7)showed a signicant increase in savings in ETwc thanks to the

Table 7

Sensitivity analysis: results are given as net difference ( D mPE/ton ww) with respect to the value of the original scenario ( D mPE = mPE new scenario-mPE original scenario).Negative values mean higher savings. Only the energy system LT-Coal (coal as marginal energy) was considered for the analysis.

Category Scenario EU MSW ENZ (; ) Parameter assessed PL REC. RR (; ) CH4 (; ) CH4 (" ) g el (" )

GW INC 8 0 0 0 0 0 0CC-CC 4 8 3 0 0 12 0CC-INC 4 8 0 0 0 6 0BG-CC 3 8 2 7 7 12 11BG-INC 2 8 0 7 7 6 11

AC INC 3 0 0 0 0 0 0CC-CC 0 0 2 0 0 9 0CC-INC 4 0 2 0 0 8 0BG-CC 3 0 5 4 4 7 2BG-INC 2 0 2 4 4 2 2

ETwc INC 25 0 0 0 0 0 0CC-CC 36 0 3 0 0 61 0CC-INC 37 0 2 0 0 21 0BG-CC 34 0 2 33 33 65 2BG-INC 34 0 2 33 33 19 2

HTw INC 18 0 0 0 0 0 0

CC-CC 33 0 1 0 0 270 0CC-INC 7 0 1 0 0 36 0BG-CC 38 0 2 0 0 268 2BG-INC 11 0 1 0 0 33 2

HTs INC 0 0 0 0 0 0 0CC-CC 2 0 0 0 0 100 0CC-INC 9 0 1 0 0 18 0BG-CC 18 0 1 0 0 98 3BG-INC 25 0 1 0 0 15 3

EU MSW = average EU municipal solid waste composition.ENZ (; ) = decreased GW impact from enzyme production (2 kg CO2-eq/kg enz).PL REC. = plastic recycling (21%).CH4 (; ) = decreased methane potential (60% VS degradation).CH4 (" ) = increased methane potential (90% VS degradation).RR (; ) = decreased recycling rate (50%).g el (" ) = higher electricity recovery in integrated gas combined cycle (60%).




higher amounts of recyclables produced. While at the same timeincreased environmental loads for the toxicity categories (HTwand HTs) were found. This was primarily because of increasedemissions of Hg, Cd, Cr, and other heavy metals to the air and soilthrough co-combustion and digestate application on soil. As shownin Table 7, the scenarios including incineration of the solid fraction(CC-INC and BG-INC) achieved a better performance in these cate-gories thanks to the improved ue-gas cleaning system (stricteremissions regulations). In the non-Toxic categories (e.g. GW andAC) the difference compared with the baseline scenario wasnegligible.

A sensitivity analysis was also performed on the impacts relatedto the industrial production of enzymes. An annual decrease of 5%on the CO2 emissions from enzymes production is expected basedon the sustainability targets of the producers ( Novozymes A/S,2004). Thus, the impact of the enzymes production was decreasedfrom 4.4 to 2 kg CO2-eq/kg enzyme (e.g. 15–20 years from the cur-rent situation). Alternatively, this corresponds to an input of 10 kg

enzymes/ton ww in the current situation (instead of 24 kg/ton ww). As shown in Table 7, the lower environmental cost of en-zymes production was crucial in amplifying the difference in theGW impact between the reference (INC) and the management sce-narios involving the waste renery. Reducing the environmentalcost of enzyme production increased the savings in GW by approx-imately 8 mPE/ton ww (i.e. 70 kg CO2-eq/ton ww).

The opportunity for plastic recycling (the recovery rate was as-sumed equal to 21%) turned out not to be signicant from an envi-ronmental point of view. The little amount of recyclable plasticfound in the waste composition (about 15 kg/ton ww) togetherwith the high savings associated with the opportunity of co-com-busting the solid fraction made the latter option preferable com-pared to recycling.

The efciency of the digestion process was found to be extre-mely relevant with respect to GW. In the sensitivity analysis, the

degradation of volatile solids (VS) was set to 60% (i.e. 67 Nm3

CH4/ton liquid fraction) and 90% (i.e. 100 Nm3 CH4/ton liquid frac-tion) (instead of 75% as it was assumed in the baseline scenario).The results (Table 7) showed a net difference compared with thebaseline scenario of about ±7 mPE/ton ww.

A lower recycling rate (assumed equal to 50% for all materials)for glass, aluminium, and ferrous metals in the waste renery ledto decreased environmental benets in the GW and ETwc catego-ries. Lower the recovery, lower the substitution of virgin resourcesand related extraction-production processes. This induced furtherenergy consumption as well as emissions contributing to impactson GW and ETwc. Furthermore, a decreased recycling rate led toan increase in the emissions of metals from the co-combustionprocess contributing to higher environmental loads in the toxicitycategories (HTs and HTw).

Finally, the biogas was assumed to be utilised in IGCC instead of

a gas-red power plant. This assumption was carried out in orderto assess the inuence of a higher electricity recovery rate on theresult. Net electricity and heat efciencies were set to 60% and30% (per LHV of the biogas) according to DEA (2005). With this op-tion, the savings in GW were increased by about 12 mPE/ton ww(i.e. 104 kg CO2-eq/ton ww).

Overall, the sensitivity analysis demonstrated that the optimi-zation of the waste renery (with downstream energy utilizations)was primarily related to metals and energy recovery (includingoptimization of biogas production) as well as the opportunity of decreasing enzymes consumption.

4. Conclusions

Four different waste management scenarios involving enzy-matic rening of residual MSW were evaluated and compared with

a reference (incineration). This was modelled across ve differentenergy systems scenarios including the substitution of coal- andnatural gas-based energy production.

The results of the study demonstrated that enzymatic reningof the waste with utilization of the products for energy recoverycan represent a valuable alternative to incineration from both anenergy and environmental point of view. This is the case if thedownstream energy options for exploiting the solid and liquid frac-tions are co-combustion and anaerobic digestion for biogas pro-duction. The principal savings of the waste renery process wererelated to higher metals and energy recovery (particularly with re-spect to electricity) compared to that of incineration. Improvementin the environmental as well as energy performance of the wasterenery itself was primarily related to the optimization of energyand enzymes consumption.

The sensitivity analysis revealed that low recycling rates formetals (under 50%) and low methane potential of the liquid frac-tion (under 70% VS degradation) would cancel the savings gained

by the waste renery, including Global Warming and Acidicationsavings from biogas production. The results also emphasized thatthe nal quality of the digestate for application on-land representsan important issue that has to be evaluated through further analy-ses and assessments.

Acknowledgements

The authors highly appreciated the contributions from NannaDreyer Nørholm and Per Lundqvist (DONG Energy) as well as theinputs from Per Henning Nielsen and Anne Merete Nielsen (Novo-zymes) regarding the production processes of enzymes. Financialsupport to this study was provided by the Research Grant PSO-7335 REnescience from Energinet.dk as well as Technical Univer-sity of Denmark.

References

Arena, U., Mastellone, M.L., Perugini, F., 2003. The environmental performance of alternative solid waste management options: a life cycle assessment study.Chemical Engineering Journal 96, 207–222.

Astrup, T., Møller, J., Fruergaard, T., 2009. Incineration and co-combustion of waste:accounting of greenhouse gases and global warming contributions. WasteManagement & Research 27, 789–799.

Birgisdottir, H., Bhander, G., Hauschild, M.Z., Christensen, T.H., 2007. Life cycleassessment of disposal of residues from municipal solid waste incineration:recycling of bottom ash in road construction or landlling in Denmarkevaluated in the ROAD-RES model. Waste Management 27, S75–S84.

Boldrin, A., Neidel, T.L., Damgaard, A., Bhander, G.S., Møller, J., Christensen, T.H.,2011. Modelling of environmental impacts from biological treatment of organicmunicipal waste in EASEWASTE. Waste Management 31, 619–630.

Cherubini, F., Ulgiati, S., 2010. Crop residues as raw materials for biorenerysystems – a LCA case study. Applied Energy 87, 47–57.

Cherubini, F., Jungmeier, G., 2010. LCA of a biorenery concept producingbioethanol, bioenergy, and chemicals from switchgrass. International Journal

of Life Cycle Assessment 15, 53–66.Cherubini, F., Stroemann, A.H., 2010. Production of biofuels and biochemicals fromlignocellulosic biomass: estimation of maximum theoretical yields andefciencies using matrix algebra. Energy & Fuels 24, 2657–2666.

Christensen, T.H., Simion, F., Tonini, D., Møller, J., 2009. Global warming factorsmodelled for 40 generic municipal waste management scenarios. WasteManagement & Research 27, 871–884.

Directive 1999/31/EC of the European Parliament and of the Council of 26 April1999 on the landll of waste. EU Commission, Brussels.

Davidsson, Å., Gruvberger, C., Christensen, T.H., Hansen, T.L., Jansen, J.C., 2007.Methane yield in source-sorted organic fraction of municipal solid waste. WasteManagement 27, 406–414.

Danish Energy Authority (DEA), 2005. Technology data for electricity and heatgenerating plants. Available from: <http://www.ens.dk >.

DONG, 2008. Avedøreværket – Grønt regnskab 2008 (Avedøreværket – Greenreport, in Danish). Available from: <http://www.dongenergy.com > (accessedMarch 2010).

DTU Environment, 2008. EASEWASTE 2008 database (Environmental Assessment of Solid Waste Systems and Technologies). http://www.easewaste.dk/ . Developedat Department of Environment Engineering, Technical University of Denmark,Kongens Lyngby, Denmark.




Eisted, R., Larsen, A.W., Christensen, T.H., 2009. Collection, transfer and transport of waste: accounting of greenhouse gases and global warming contribution. WasteManagement & Research 27, 738–745.

Ekvall, T., Weidema, B.P., 2004. System boundaries and input data in consequentiallife cycle inventory analysis. International Journal of Life Cycle Assessment 9(3), 161–171.

Finnveden, G., Hauschild, M.Z., Ekvall, T., Guine, J., Heijungs, R., Hellweg, S., Koehler,A., Pennington, D., Suh, S., 2009. Recent developments in Life Cycle Assessment. Journal of Environmental Management 91, 1–21.

Fruergaard, T., Astrup, T., Ekvall, T., 2009. Energy use and recovery in wastemanagement and implications for accounting of greenhouse gases and globalwarming contributions. Waste Management Research 27, 724–737.

Fruergaard, T., Christensen, T.H., Astrup, T., 2010a. Energy recovery from wasteincineration: assessing the importance of district heating networks. WasteManagement 30, 1264–1272.

Fruergaard, T., Hyks, J., Astrup, T., 2010b. Life-cycle assessment of selectedmanagement options for air pollution control residues from wasteincineration. Science of the Total Environment 408, 4672–4680.

Fruergaard, T., Astrup, T., 2011. Optimal utilization of waste-to-energy in an LCAperspective. Waste Management 31, 572–582.

Hansen, E., Olsen, S.I., Schmidt, A., Hauschild, M., Hjelmar, O., Bendtsen, N., Poulsen,T.S., Hansen, H.H., Christensen, K., 2004. Life cycle assessment of landlledwaste (in Danish). Environmental Project No. 971. Danish EnvironmentalProtection Agency, Copenhagen, Denmark.

Jensen, J.W., Felby, C., Jørgensen, H., Rønnsch, G.Ø., Nørholm, N.D., 2010. Enzymaticprocessing of municipal solid waste. Waste Management.Kaparaju, P., Serrano, M., Thomsen, A.B., Kongjan, P., Angelidaki, I., 2009. Bioethanol,

biohydrogen and biogas production from wheat straw in a biorenery concept.Bioresource Technology 100, 2562–2568.

Larsen, A.W., Merrild, H., Møller, J., Christensen, T.H., 2010. Waste collectionsystems for recyclables: an environmental and economic assessment for themunicipality of Aarhus (Denmark). Waste Management 30, 744–754.

Larsen, J., Petersen, M.Ø., Thirup, L., Li, H.W., Iversen, F.K., 2008. The IBUS process –lignocellulosic bioethanol close to a commercial reality. Chemical Engineering &Technology 31, 765–772.

Lohrasby, M., Pourbafrani, M., Niklasson, C., Taherzadeh, J., 2010. Process design andeconomic analysis of a citrus waste biorenery with biofuels and limonene asproducts. Bioresource Technology 101, 7382–7388.

Mao, H., Genco, J.M., van Heiningen, A., Pendse, H., 2010. Kraft mill biorenery toproduce acetic acid and ethanol: technical economic analysis. BioresourceTechnology 5, 525–544.

Møller, J., Jansen, J.L., Cour, Christensen, T.H., 2010. Anaerobic digestion: massbalances and products. Solid Waste Technology and Management 2, 618–629.

Nielsen, P.H., Oxenbll, K.M., Wenzel, H., 2007. Cradle-to-gate environmentalassessment of enzyme products produced industrially in Denmark byNovozymes A/S. Int J LCA 12 (6), 432–438 (Updated values (2010)).

Novozymes A/S, 2004. Report. America.Nielsen, M., Illerup, J.B., 2006. Danish emission inventories for stationary

combustion plants. Inventories until year 2003. Research Notes NERI No. 229.National Environmental Research Institute, Denmark.

Papatheofanous, M.G., Koullas, D.P., Koukios, E.G., Fuglsang, H., Schade, J.R., Löfqvist,B., 1995. Biorening of agricultural crops and residues: effect of pilot-plantfractionation on properties of brous fractions. Biomass & Bioenergy 8, 419–426.

Pognani, M., D’Imporzano, G., Scaglia, B., Adani, F., 2009. Substituting energy cropswith organic fraction of municipal solid waste for biogas production at farmlevel: a full-scale plant study. Process Biochemistry 44, 817–821.

Riber, C., Petersen, C., Christensen, T.H., 2009. Chemical composition of materialfractions in Danish household waste. Waste Management 29, 1251–1257.

Rigamonti, L., Grosso, M., 2009. Inuence of assumptions about selection andrecycling efciencies on the LCA of integrated waste management systems. Int JLCA 411, 419.

Sander, B., Christensen, B.T., Hansen, E.M., Magid, J., Jensen, L.S., 2003. Calibrationand validation of the soil organic matter dynamics of the Daisy model with datafrom the Askov long-term experiments. Soil Biology and Biochemistry 35, 67–76.

Schmidt, A., Stromberg, K., 2006. Genanvendelse i LCA – systemudvidelse (Recycling

in LCA – system expansion, in Danish), Miljoenyt Nr. 81. Danish EnvironmentalProtection Agency, Copenhagen, Denmark.Stranddorf, H.K., Hoffmann, L., Schmidt, A., 2005. Påvirkningskategorier,

normalisering og vægtning i LCA – opdatering af udvalgte UMIP97-data.Danish Ministry of the Environment, Miljønyt n. 77.

Talebnia, F., Karakashev, D., Angelidaki, I., 2010. Production of bioethanol fromwheat straw: an overview on pretreatment, hydrolysis and fermentation.Bioresource Technology 101 (13), 4744–4753.

Uihlein, A., Schebek, L., 2009. Environmental impacts of a lignocellulose feedstockbiorenery system: an assessment. Biomass and Bioenergy 33, 793–802.

Villegas, J., Gnansounou, E., 2008. Techno-economic and environmental evaluationof lignocellulosic biochemical reneries: need for a modular platform forintegrated assessment (MPIA). Journal of Scientic & Industrial Research 67,927–940.

Weidema, B., Frees, N., Nielsen, A.M., 1999. Marginal production technologiesfor Life Cycle Inventories. International Journal of Life cycle Assessment 4,48–56.

Wenzel, H., Hauschild, M.Z., Alting L., 1997. Environmental Assessment of Products,vol. 1.




IV

Advanced material, substance and energyflow analysis of a waste refinery process

Tonini, D., Dorini, G., Astrup, T.

Bioresource Technology, submitted.



Advanced material, substance and energyflow analysis of a waste refinery process

Davide Tonini 1*, Gianluca Dorini 2, Thomas Astrup 3

1, 3 Department of Environmental Engineering, Technical University of Denmark, DTU,Miljoevej, Building 115, 2800 Kgs. Lyngby, Denmark

2 Department of Informatics and Mathematical Modelling, Technical University ofDenmark, DTU, Building 305, 2800 Kgs. Lyngby, Denmark

* Corresponding author: [email protected] 0045 45251699



1

Abstract

Materials and energy recovery from mixed household waste may contribute to decrease

fossil fuel and resource consumption. To this purpose, legislations have been enforced

to limit landfilling and promote energy and materials recovery. Potential solutions in

separating degradable and recyclable materials are offered by the waste refineries where

a bioliquid is produced from enzymatic treatment of mixed waste. In this study,

potential flows of materials, energy and substances within a waste refinery are

investigated by combining sampling, analyses and modeling. The results highlighted

that the waste refinery may recover ca. 56% of the input dry matter as bioliquid yielding

6.2 GJ biogas-energy. The potential for N, P, K and biogenic carbon recovery was

estimated between 83% and 93% of the input. Additional metals and plastic recovery is

possible. A drawback may be the digestate quality, which may not comply with selected

use on land criteria.

Keywords : Waste refinery, MFA, SFA, EFA, Waste sampling, Waste characterization

1. Introduction

Recovery of materials, resources and energy from different waste types has become a

crucial aspect of waste management strategies in the light of the new European

framework directive on waste management (The European Parliament and The Council,

2008). The target of the directive is to minimize the amount of waste disposed of in

landfills (especially with respect to organic) and to optimize instead the recovery of

valuable materials (e.g. metals, paper and plastic), resources (e.g. phosphorous from

organics) and energy. Although technologies exist for sorting of selected waste material

fractions, an efficient separation of organic materials for energy and nutrients recovery

and of recyclables to reduce resource consumption is important, but difficult on mixed

household waste. In addition, in the regions where landfilling (instead of, for example,

incineration) is the most common disposal method, separation of the degradable

organics (e.g. kitchen waste, tissues, etc.) is a priority in order to comply with the

reduction targets established by the European Union (CEC, 1999). Organic waste

source-segregation at the household may contribute to this goal; however, recent studies

have highlighted that such a strategy may end up being inefficient (mass- and energy-



3

suggest system improvements on the basis of the results. Further, the results of MFA

and SFA are often used as basis for life cycle assessment (LCA). In this perspective,

MFA, SFA and LCA represent complementary tools for environmental management

(Pires et al., 2011). For instance, Mastellone et al. (2009) used MFA to identify the

relevant waste flows in a waste-emergency area and to suggest management solutions;

Andersen et al. (2010) combined MFA and LCA to assess the performance of a garden

waste composting plant; Gurauskienė and Stasiškienė (2011) and Oguchi et al. (2012)

used MFA and SFA to estimate flows and recycling efficiencies for electronic waste;

Pomberger et al. (2012) modelled the energy content of solid recovered fuel (SRF)

based on MFA. Tonini et al. (2012) combined SFA and LCA to assess the performance

of bioenergy scenarios. However, in addition to mass and substance flows, in order to

address the theoretical performance of pilot-scale waste refineries, mathematical

modeling needs to be applied for determining the potential optimum recovery of

bioliquid, materials and nutrients thereby providing a target for further technological

development.

This study used an advanced MFA, SFA and EFA approach based on a

mathematical optimization model to evaluate the potential flows of materials,

substances (e.g. carbon, nutrients and selected metals) and energy within a wasterefinery including downstream energy conversion processes. The objectives of the study

were: I) a detailed sampling and characterization of the outputs of a pilot-scale waste

refinery process (materials flow and chemical composition) with particular focus on the

bioliquid; II) the development of a mathematical optimization model to evaluate the

potential for recovery of bioliquid, materials and nutrients with a ‘virtual’ post-

treatment phase; III) the development of MFA, SFA and EFA models based on the

mathematical model outputs to illustrate the potential flows of materials, energy, carbon(including carbon fossil), nutrients and selected metals (Al and Fe); IV) the evaluation

of the quality of the digestate left after anaerobic digestion of the bioliquid in order to

assess the load of nutrients and metals in the scenario of application on land.



4

2. Materials and methods

The study involved 5 major phases: 1) on-field sampling of the pilot-scale waste

refinery outputs (bioliquid, fluff and solid fraction ex-enzymatic treatment); hand-

sorting of the solid fraction was also performed at this point: 6 individual waste material

fractions were sorted and separated (see 2.2). Overall, 8 waste material fractions were

thus collected (6 from the solid fraction ex-enzymatic treatment plus bioliquid and

fluff). 2) Preparation of the 8 individual samples for chemical analyses (shredding,

mixing, splitting, etc.). 3) Chemical composition analyses (including calorific value). 4)

Elaboration of a mathematical optimization model to estimate the potential for

bioliquid, materials and nutrients recovery with a ‘virtual’ post-treatment. 5)

Elaboration of MFA, SFA and EFA to illustrate material, substance and energy flows

within the waste refinery process including virtual post-treatment and downstream

energy and materials recovery processes. Table 1 summarizes the 5 phases of the study

with the associated methods applied.



6

2.1 The waste refinery process

The study was based on the operation of a pilot-scale plant (0.5-1 Mg wet waste (ww) h -

1) where the waste was processed (heating and enzymatic treatment) without further

post-treatment. The pilot-scale plant treated residual municipal solid waste (rMSW)

collected from a residential district of Copenhagen (Denmark) where a vacuum-

collection system is established. The waste was sampled and characterized within this

study (as output of the waste refining process, see section 3).

The waste refinery aims at producing two products from the incoming MSW: i)

a bioliquid (i.e. slurry composed of enzymatically liquefied organic, paper and

cardboard) and a solid fraction (i.e. non-degradable waste materials). The refinery

process consisted of two reactors: in the first reactor the waste was heated to about 75

°C for approximately 0.5-1 hour, then cooled to about 50-55 °C before entering the

second reactor. In the second reactor selected enzymes were added (ca. 5 kg) resulting

in hydrolysis and break-down of bonds in the organic materials thereby essentially

suspending organic materials in a liquid phase. The retention time was about 10-16

hour. A detailed description of the enzymatic processing can be found in Jensen et al.

(2010). After the second reactor, the liquid phase was separated from the remaining

solid fraction by a vibrating sieve. Later, another vibrating sieve separated the liquid phase into a bioliquid and a solid “fluff” (phase containing materials such as cotton and

textiles, but also glass pieces, plastics, etc.). The bioliquid consisted primarily of

suspended organic matter (food waste, paper and cardboard), while the solid fraction

mainly consisted of non-degradable materials such as plastic, metals, textiles, soil,

ceramics, glass pieces, mixed with not separated bioliquid. The solid fraction and the

fluff require post-treatment to recover additional bioliquid (through for example

washing and pressing). The bioliquid can be exploited for biogas production (option considered in this

study), co-combusted in coal-fired power plant or utilized for producing ethanol (Tonini

and Astrup, 2012). Biogas production, as compared with direct incineration, provides

additional flexibility to the energy system as the energy production could be regulated

and storage is possible. This is crucial in the perspective of energy systems with high

penetration of wind and other fluctuating renewables as illustrated in previous studies

(Lund and Mathiesen, 2009; Tonini and Astrup, 2012). The solid fraction can undergo



9

ARP CS 2000 shredder, pressed into cubes, mass-reduced by titanium

drilling and the obtained scraps were further mass-reduced with the disc-mill

Siebtechnik IS100A, Mülheim an der Ruhr, down to a size of less than 1 x 1

mm. This laborious method was chosen because these fractions could not be

directly cut, shredded or crushed due to their strength.

• Dried residue (RES*) was shredded with the Retsch SM 2000 cutter mill

down to a size of less than 1 x 1 mm.

• Dried fluff (FF*) was shredded with the Retsch SM 2000 cutter mill down to

a size of less than 1 x 1 mm.

During all operations large amounts of pulverized dry ice (CO 2,S) were added during

handling to ensure sufficient cooling capacity and to facilitate the shredding/milling.

Mixing and fractional mass reduction were done in the same way for all fractions by

repeated mixing in a mechanical mixer (or by hand) and then mass-reduced with a riffle

splitter (Rationel Kornservice RK12, Esbjerg, Denmark) until the required mass for

chemical analysis was obtained.

2.4 Chemical composition analyses (phase 3)

Volatile solid (VS) content, chemical composition, calorific value (higher heating value,HHV) was determined by standard analyses following the approach described in Riber

et al. (2007) specific for solid waste material fractions. Volatile solid content was

measured by incinerating the dried grinded samples at 550 ºC in muffle. The substances:

C, H, N, S were analyzed with elemental analysis (GE-MA M-7-1); Chloride and F

were analyzed conformingly with DIN 51727 B; phosphorous was analyzed with

inductively coupled plasma optical emission spectrometry (ICP-OES) conformingly

with DIN EN ISO 11885 (E22). The remaining metals: Fe, Al, As, Cd, Cr, Cu, Fe, Hg,Mn, Mg, Ni, Pb, Sb and Sr were analyzed with ICP-MS conformingly with DIN EN

ISO 17294-2 (E29). The HHV db (dry basis) was determined with calorimetric bomb

conformingly with DIN 51900. The content of biogenic carbon (represented by the 14C

content in 12C) was analyzed with accelerated mass spectrometry (AMS) conformingly

with CEN/TS 15747:2008. The theoretical methane potential of the bioliquid (B o,th) was

estimated by applying the Buswell’s formula (Eq. 1) provided the composition of the

bioliquid in terms of carbohydrates, proteins, lipids, volatile fatty acids (VFA) and



10

ethanol (EtOH) determined with high performance liquid chromatography (HPLC) . In

addition, batch assays were also conducted according to the procedure described in

Angelidaki et al. (2009). The batch assays were conducted in 500 mL serum bottles

inoculated with digested manure from termophilic digestion plants and fresh bioliquid.

Tests were run in triplicates for 52 days at 55 ºC (Golisowicz, 2011).

EtOH VFAlipids. proteins+.tes+carbohydra= B th ⋅+⋅+⋅⋅⋅ 73.0373.001414960415.0,0

(1)

2.4.1 Chemical composition of the digestate

The composition of the digestate was determined through i) SFA modelling and ii)

chemical analyses on actual samples from full-scale anaerobic digestion of the

bioliquid. The first approach was based on the chemical composition of the sampled

bioliquid (BL*) assuming that heavy metals (as well as P and K) were entirely

transferred to the digestate and modeling expected dry matter degradation occurring

during the digestion (see section 2.6 and supporting information, SI); the second

approach provided instead a ‘snap-shot’ of the chemical composition based on actual

digestate samples representing a specific point in time. Selected parameters and

substances (e.g. DM, VS, C, N, P, K, metals, etc.) were analyzed. The substances: C, N,

H were analyzed with elemental analysis (GE-MA M-7-1). The metals (including P and

K) were analyzed with ICP OES conformingly with EN 13346:2000. The measured VS,

C and N content were used to calculate the expected DM, C and N degradation

occurring during the anaerobic digestion process (see section 2.6 and SI). In addition,

the analyses on the metals provided a set of concentration values that were compared

with the SFA modeling results and further used as basis to evaluate the quality of thedigestate in the scenario of application on land (see section 3.4).

http://www.youtube.com/watch?v=kz_egMtdnL4

http://www.youtube.com/watch?v=kz_egMtdnL4



12

Fig. 2 . Illustration of the principles of the optimization model. Nomenclature as in Fig. 1.

2.5.1 Formulation of the mathematical model

The mathematical formulation consisted of two parts: i) mass balance and ii) uncertainty

handling. The first models how to break down the total dry mass flow into bioliquid and

material flow, and how to determine the chemicals concentration within the material

flow. Within this part, I) the total (dry) waste materials flow, II) the total chemicalsflow, and III) the bioliquid chemicals flow, are known (from sampling and analyses),

i.e. they are input to the mathematical model (Table 1). First, the chemical

concentrations in the material flow are determined by solving, for each flow, a simple

optimization problem. Then, the bioliquid and material component of the flow is

calculated immediately from the chemicals.

The second part uses Probability Distribution Functions to characterize the

uncertainty in the model input variables and to propagate such uncertainty onto theresults. A detailed description of the mathematical formulation is reported in the

supporting information (SI) (an overview is also presented in Fig. S1 of the SI).

2.6 Material, substance and energy flow analysis (phase 5)

The results of the mathematical optimization model ( BLi, M i and CM ji) were used as

inventory for the MFA, SFA and EFA of the waste refinery process. This was facilitated

by the software STAN (Cencic and Rechberger, 2008). The MFA encompasses the



1 5

T a

b l e 2

. C h e m i c a l c o m p o s i t i o n o f

t h e s a m p l e d

( * ) a n d m o d e l e d w a s t e m a t e r i a l

f r a c t i o n s ( k g k g - 1 D M ; v a l u e s r o u n d e d

t o 2 s i g n i f i c a n t d i g i t s ) .

B L : b i o l i q u i d ;

F E : f e r r o u s m a t e r i a l ; F F : f l u f f ; H

P : h a r d p l a s t i c ; N

F E : n o n - f e r r o u s m a t e r i a l ; R

E S : s o l i d r e s i d u e ; S P : s o f t p l a s t i c ; T

X T : t e x t i l e s ; T / R : r a t i o

b e t w e e n

t h e t o t a l

a m o u n t o f

t h e s e l e c t e d c h e m

i c a l i n t h i s s t u d y a n d

i n R i b e r e t a l . ( 2 0 0 9 ) ; μ : s h a r e

( a s % ) o f

t h e t o t a l D M

. σ : s t a n d a r d d e v i a t i o n o f μ

( e . g .

t h e s h a r e o f

H P * o n

t h e t o t a l D M i s 1 0 % ± 2 . 8 % ) . O & : s u m o f o x y g e n a n d

( a g g r e g a t e d ) h e a v y m e t a l s

i n t r a c e s ( s e e s e c t i o n

2 . 5 . 1 ) ; a g g . : a g g r e g a t e d i n

O & ; n . r . : n o t r e p o r t e d .

B L * ( B L )

H P *

H P

S P *

S P

T X T *

T X T

F E *

F E

N F E *

N F E

R E S *

R E S

F F *

F F

T O T A L

T / R

μ

9 . 0 E - 0 2 α

1 . 0 E - 0 1

8 . 2 E - 0 2

2 . 1 E - 0 1

9 . 0 E - 0 2

1 . 7 E - 0 1

7 . 6 E - 0 2

5 . 0 E - 0 2

4 . 1 E - 0 2

2 . 8 E - 0 2

2 . 4 E - 0 2

3 . 5 E - 0 1

1 . 2 E - 0 1

9 . 1 E - 0 3

4 . 0 E - 0 3

1 . 0 E + 0 0

1 . 0

σ

4 . 1 E - 0 2

2 . 8 E - 0 2

2 . 3 E - 0 2

7 . 6 E - 0 2

3 . 4 E - 0 2

4 . 4 E - 0 2

2 . 1 E - 0 2

2 . 8 E - 0 2

2 . 3 E - 0 2

1 . 4 E - 0 2

1 . 2 E - 0 2

1 . 3 E - 0 1

4 . 7 E - 0 2

4 . 1 E - 0 4

2 . 2 E - 0 4

-

n . r .

C b i o g

4 . 2 E - 0 1

9 . 4 E - 0 2

2 . 0 E - 0 2

3 . 0 E - 0 1

1 . 4 E - 0 1

4 . 0 E - 0 1

3 . 7 E - 0 1

8 . 7 E - 0 2

1 . 7 E - 0 2

1 . 2 E - 0 1

2 . 0 E - 0 5

3 . 1 E - 0 1

1 . 1 E - 0 1

3 . 7 E - 0 1

3 . 0 E - 0 1

2 . 9 E - 0 1

1 . 0

C f o s s

8 . 3 E - 0 3

6 . 4 E - 0 1

7 . 8 E - 0 1

2 . 4 E - 0 1

5 . 4 E - 0 1

3 . 9 E - 0 2

7 . 6 E - 0 2

5 . 3 E - 0 2

6 . 2 E - 0 2

2 . 9 E - 0 2

1 . 8 E - 0 2

1 . 2 E - 0 1

3 . 2 E - 0 1

7 . 0 E - 0 2

1 . 5 E - 0 1

1 . 7 E - 0 1

1 . 3

H

5 . 1 E - 0 2

1 . 1 E - 0 1

1 . 2 E - 0 1

8 . 1 E - 0 2

1 . 2 E - 0 1

6 . 2 E - 0 2

7 . 5 E - 0 2

4 . 5 E - 0 2

4 . 4 E - 0 2

3 . 2 E - 0 2

2 . 8 E - 0 2

6 . 0 E - 0 2

7 . 6 E - 0 2

6 . 7 E - 0 2

8 . 7 E - 0 2

6 . 7 E - 0 2

1 . 1

S

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

1 . 0 E - 0 3

0 . 7 2

N

2 . 0 E - 0 2

3 . 7 E - 0 3

2 . 3 E - 0 6

1 . 2 E - 0 2

1 . 9 E - 0 3

1 . 1 E - 0 2

3 . 9 E - 0 4

5 . 0 E - 0 3

1 . 9 E - 0 3

2 . 3 E - 0 2

2 . 4 E - 0 2

1 . 4 E - 0 2

3 . 1 E - 0 3

3 . 6 E - 0 2

5 . 6 E - 0 2

1 . 3 E - 0 2

1 . 0

F

1 . 8 E - 0 4

7 . 0 E - 0 5

4 . 5 E - 0 5

1 . 2 E - 0 4

4 . 4 E - 0 5

1 . 1 E - 0 4

2 . 8 E - 0 5

9 . 0 E - 0 5

7 . 1 E - 0 5

1 . 2 E - 0 4

1 . 1 E - 0 4

1 . 2 E - 0 4

1 . 1 E - 0 5

1 . 5 E - 0 4

1 . 1 E - 0 4

1 . 2 E - 0 4

0 . 9

P

2 . 5 E - 0 3

4 . 8 E - 0 4

2 . 2 E - 0 5

1 . 4 E - 0 3

1 . 1 E - 0 5

1 . 7 E - 0 3

7 . 6 E - 0 4

5 . 1 E - 0 4

9 . 4 E - 0 5

6 . 9 E - 0 4

3 . 5 E - 0 4

2 . 4 E - 0 3

2 . 2 E - 0 3

1 . 9 E - 0 3

1 . 2 E - 0 3

1 . 7 E - 0 3

0 . 8 4

C l

1 . 1 E - 0 2

2 . 2 E - 0 3

2 . 1 E - 0 4

1 . 6 E - 0 2

2 . 2 E - 0 2

6 . 0 E - 0 3

1 . 4 E - 0 4

1 . 9 E - 0 3

8 . 9 E - 0 7

2 . 2 E - 0 3

5 . 3 E - 0 4

8 . 5 E - 0 3

4 . 0 E - 0 3

6 . 1 E - 0 3

1 . 5 E - 0 6

8 . 7 E - 0 3

1 . 0

K

7 . 0 E - 0 3

1 . 5 E - 0 3

2 . 5 E - 0 4

5 . 0 E - 0 3

2 . 5 E - 0 3

5 . 1 E - 0 3

2 . 9 E - 0 3

1 . 3 E - 0 3

1 . 3 E - 0 4

1 . 7 E - 0 3

7 . 0 E - 0 4

4 . 5 E - 0 3

4 . 7 E - 0 5

6 . 0 E - 0 3

4 . 7 E - 0 3

4 . 4 E - 0 3

1 . 0

F e

1 . 7 E - 0 3

1 . 7 E - 0 3

1 . 7 E - 0 3

2 . 1 E - 0 3

2 . 5 E - 0 3

1 . 9 E - 0 3

2 . 2 E - 0 3

6 . 7 E - 0 1

8 . 1 E - 0 1

1 . 1 E - 0 1

1 . 3 E - 0 1

3 . 4 E - 0 3

6 . 5 E - 0 3

3 . 8 E - 0 3

6 . 4 E - 0 3

3 . 9 E - 0 2

1 . 4

A l

3 . 0 E - 0 3

1 . 2 E - 0 3

8 . 1 E - 0 4

2 . 1 E - 0 3

8 . 8 E - 0 4

2 . 4 E - 0 3

1 . 7 E - 0 3

2 . 3 E - 0 2

2 . 8 E - 0 2

6 . 0 E - 0 1

7 . 1 E - 0 1

7 . 0 E - 0 3

1 . 4 E - 0 2

3 . 7 E - 0 3

4 . 6 E - 0 3

2 . 2 E - 0 2

0 . 9 4

C d

1 . 8 E - 0 7

1 . 0 E - 0 7

a g g .

1 . 1 E - 0 7

a g g .

1 . 7 E - 0 7

a g g .

1 . 0 E - 0 7

a g g .

1 . 1 E - 0 7

a g g .

1 . 1 E - 0 7

a g g .

1 . 3 E - 0 7

a g g .

1 . 2 E - 0 7

0 . 0 2

C r

1 . 9 E - 0 5

1 . 3 E - 0 5

a g g .

1 . 3 E - 0 5

a g g .

8 . 3 E - 0 6

a g g .

2 . 3 E - 0 4

a g g .

1 . 7 E - 0 4

a g g .

1 . 7 E - 0 5

a g g .

1 . 2 E - 0 5

a g g .

2 . 9 E - 0 5

0 . 3

C u

3 . 5 E - 0 5

1 . 2 E - 0 5

a g g .

7 . 5 E - 0 5

a g g .

3 . 6 E - 0 5

a g g .

1 . 3 E - 0 4

a g g .

6 . 7 E - 0 4

a g g .

4 . 8 E - 0 5

a g g .

2 . 7 E - 0 5

a g g .

6 . 8 E - 0 5

0 . 1

N i

1 . 2 E - 0 5

5 . 7 E - 0 6

a g g .

4 . 7 E - 0 6

a g g .

4 . 2 E - 0 6

a g g .

3 . 3 E - 0 4

a g g .

2 . 2 E - 0 4

a g g .

2 . 4 E - 0 5

a g g .

5 . 3 E - 0 6

a g g .

3 . 4 E - 0 5

0 . 7 1

S r

1 . 0 E - 0 4

2 . 0 E - 0 5

a g g .

6 . 0 E - 0 5

a g g .

6 . 2 E - 0 5

a g g .

1 . 2 E - 0 4

a g g .

2 . 1 E - 0 5

a g g .

6 . 7 E - 0 5

a g g .

1 . 0 E - 0 4

a g g .

6 . 5 E - 0 5

n . r .

M n

8 . 0 E - 0 5

3 . 4 E - 0 5

a g g .

4 . 6 E - 0 5

a g g .

6 . 4 E - 0 5

a g g .

1 . 9 E - 0 3

a g g .

5 . 4 E - 0 3

a g g .

5 . 6 E - 0 3

a g g .

7 . 3 E - 0 5

a g g .

2 . 3 E - 0 3

6 . 9

M g

2 . 2 E - 0 3

7 . 1 E - 0 4

3 . 6 E - 0 4

1 . 3 E - 0 3

5 . 7 E - 0 5

1 . 3 E - 0 3

1 . 6 E - 0 4

1 . 3 E - 0 3

1 . 1 E - 0 3

9 . 6 E - 0 3

1 . 1 E - 0 2

1 . 8 E - 0 3

1 . 0 E - 0 3

2 . 5 E - 0 3

2 . 8 E - 0 3

1 . 7 E - 0 3

1 . 1

A s

1 . 4 E - 0 6

1 . 0 E - 0 6

a g g .

1 . 0 E - 0 6

a g g .

1 . 0 E - 0 6

a g g .

1 . 1 E - 0 5

a g g .

2 . 5 E - 0 6

a g g .

1 . 0 E - 0 6

a g g .

1 . 4 E - 0 6

a g g .

1 . 6 E - 0 6

0 . 2

H g

1 . 0 E - 0 7

1 . 0 E - 0 7

a g g .

1 . 0 E - 0 7

a g g .

1 . 0 E - 0 7

a g g .

1 . 0 E - 0 7

a g g .

1 . 0 E - 0 7

a g g .

1 . 0 E - 0 7

a g g .

5 . 0 E - 0 7

a g g .

1 . 0 E - 0 7

0 . 0 4

P b

1 . 2 E - 0 5

2 . 2 E - 0 5

a g g .

7 . 6 E - 0 6

a g g .

8 . 0 E - 0 6

a g g .

4 . 7 E - 0 6

a g g .

1 . 3 E - 0 5

a g g .

1 . 2 E - 0 5

a g g .

2 . 3 E - 0 4

a g g .

1 . 3 E - 0 5

0 . 0 6

S b

1 . 4 E - 0 6

3 . 5 E - 0 5

a g g .

1 . 0 E - 0 6

a g g .

6 . 7 E - 0 6

a g g .

3 . 8 E - 0 6

a g g .

1 . 0 E - 0 6

a g g .

1 . 5 E - 0 6

a g g .

1 . 0 E - 0 6

a g g .

5 . 7 E - 0 6

n . r .

O &

4 . 7 E - 0 1

1 . 4 E - 0 1

6 . 9 E - 0 2

3 . 4 E - 0 1

1 . 7 E - 0 1

4 . 7 E - 0 1

4 . 7 E - 0 1

1 . 1 E - 0 1

3 . 8 E - 0 2

6 . 8 E - 0 2

7 . 0 E - 0 2

4 . 6 E - 0 1

4 . 6 E - 0 1

4 . 3 E - 0 1

3 . 8 E - 0 1

3 . 8 E - 0 1

n . r .

H H V

d b

1 7

3 3

3 7

2 3

3 1

1 8

1 9

1 1

9 . 3

2 5

2 6

1 7

1 8

1 5

1 3

2 0

1 . 1

L H V

d b

1 6

3 1

3 4

2 1

2 8

1 7

1 8

9 . 6

8 . 3

2 4

2 6

1 6

1 6

1 4

1 1

1 9

1 . 0

F o r t h e s a m p l e d

b i o l i q u i d ( B L * ) , t h e D M

c o n t e n t w a s

2 0 % . T h e

V S w a s

8 5 % ( o f D M ) . C a r b o h y d r a t e s ,

l i p i d s , p r o t e i n s ,

V F A a n d e t h a n o l w e r e

( % V S ) : 8 5 , 2 . 8 , 1 . 9 a n d

2 . 2 .

α T h e s h a r e o f

B L o n

t h e t o t a l D M

e q u a l e d 5 . 6 E - 0 1 a f t e r s u m m

i n g t h e i n d i v i d u a l c o n t r i b u t i o n s f r o m

t h e r e m a i n i n g w a s t e m a t e r i a l f r a c t i o n s u n d e r g o i n g p o s

t - t r e a t m e n t ( F i g s .

2 - 3 ) .



16

3.2 Material and energy flows

The material and energy flows are reported in Fig. 3. Uncertainties related to relevant

flows are mentioned in brackets (e.g. ± ‘value’) as standard deviation of the mean value. Notice that the values reported in the text are expressed per 1 Mg dw (dry weight),

unless differently specified.

Fig. 3 . Mass and energy flows within the waste refinery and downstream energy conversion

processes (kg DM and GJ). MSW: municipal solid waste input; MSW’: MSW after heating andenzymatic treatment; APC: air pollution control residue; BA: bottom ash; BG: biogas; BL:

bioliquid; DG: digestate; FE: ferrous material; FF: fluff; FG: flue gas; HP: hard plastic; LF:liquid fraction; NFE: non-ferrous material; RES: residue; SP: soft plastic; TXT: textiles;*sampled waste material fractions



18

compared with the scenario involving hard plastic separation. However, incineration of

this fraction would also induce additional direct CO 2 emissions at the stack (see section

3.3).

3.3 Substance flows

The substance flows for C, C foss , N, P, K, Fe and Al are illustrated in Figs. 4-6. These

are the most important substances with respect to biogas (C), nutrients (N, P and K) and

materials recovery (Fe, Al and C foss for plastic). Heavy metals flows (e.g. Cd, Ni, Pb,

Cu, etc.) are also important for the use of land of the digestate left from anaerobic

digestion. To this respect, ad hoc characterization of the digestate was performed (see

section 3.4). Notice that the values reported in the text are expressed per 1 Mg dw,unless differently specified.

Most of the carbon (Fig. 4) was found in the bioliquid flow (240 kg) of which

almost 100% was biogenic carbon (Fig. 4). Based on the mass balances detailed in the

SI (Eq. S11-S14) about 75% of this carbon was biogasified during anaerobic digestion

which reduced the carbon left in the digestate to about 60 kg. Significant fossil carbon

flows (Fig. 4) were embedded into hard plastic (65 kg), soft plastic (49 kg) and residue

(40 kg). Overall, under the assumption that hard plastic was separated and recovered,‘residual’ 95 kg of fossil carbon were incinerated corresponding to a direct emission of

ca. 350 kg CO 2. In other words, sorting and recovery of hard plastic (HP) avoided ca.

240 kg direct CO 2 emission. Additional direct CO 2 emission savings may be achieved

by separating and recovering soft plastic (potential 180 kg CO 2) and eventually other

plastic items found in the residue.

About 85% of the nitrogen (i.e. 11 kg) was found in the bioliquid flow (Fig. 5)

of which 87% (i.e. 9.7 kg) ended up in the digestate. A minor amount of nitrogen was

also found in the residue, textile and soft plastic sent to incineration (overall ca. 0.8 kg)

and in the biogas (1.4 kg). The phosphorous (Fig. 5) was primarily collected in the

bioliquid flow (about 1.4 kg, i.e. 83% of the total) although a share was incinerated

through textiles, residue and soft plastic (overall 0.33 kg). As for N and P, also the K

(Fig. 5) was recovered mainly in the bioliquid flow (about 3.9 kg, i.e. 93% of the total).

As expected, almost all the iron (Fe) and aluminium (Al) was found, respectively, in the

ferrous (about 33 kg, i.e. 80% of the total) and non-ferrous (about 19 kg, i.e. 77% of the



23

Table 3 . Modeled and analyzed concentrations of selected chemicals in the digestate from bioliquid digestion and in selected digests from source-separated organic waste processing. Thevalues (rounded to 2 significant digits) are expressed as mg kg -1 DM (except for DM, VS, C and

N which are expressed as %DM). * Dewatered sample. (1) and (2): Møller et al. (2010); (3):Holm et al. (2010); (4a) and (4b): concentration limit for use on-land (mg kg -1 DM and mg kg -1 P, respectively) as reported in Danish Ministry of Food, Agriculture and Fisheries (2006); DG:digestate; n.d.: not detected; n.m.: not modeled; n.a.: not analyzed; n.r.: not reported. PAH: sumof polyaromatic hydrocarbons; DEHP: diethylhexyl phthalate; LAS: linear alkylbenzenesulfonate; NPs: sum of nonylphenols.Chemical DG (model) DG (analysis) DG (analysis)* (1) (2) (3) (4a) (4b)

DM n.m. 4.2 ±0.1 16 44 2.7 1-2 n.r. n.r.

VS n.m. 59 59 n.r. n.r. 65 n.r. n.r.

C 28 ±8 28 29 n.r. n.r. 39 n.r. n.r.

H n.m. n.a. n.a. n.r. n.r. n.r. n.r. n.r.

S 2600 ±700 5600 6000 n.r. n.r. 5000 n.r. n.r.

N 4.6 ±1.2 4.6 ±0.14 n.a. n.r. n.r. 3.5 n.r. n.r.

F 470 ±130 n.a. n.a. n.r. n.r. n.r. n.r. n.r.P 6600 ±1700 6400 ±900 6000 n.r. 7963 9000 n.r. n.r.

Cl 29000 ±7600 n.a. n.a. n.r. n.r. n.r. n.r. n.r.

K 18000 ±5000 21000 7800 n.r. n.r. n.r. n.r. n.r.

Fe 13000 ±360 17000 20000 n.r. n.r. n.r. n.r. n.r.

Al 8000 ±2200 8100 11000 n.r. n.r. n.r. n.r. n.r.

Cd 0.47 ±0.12 ε 0.75 ±0.029 α 0.70 β 0.76 0.95 0.3-0.7 0.8 100

Cr 50 ±13 53 ±2.5 57 n.r. 9.9 n.r. 100 n.r.

Cu 92 ±24 120 ±5.9 130 n.r. 76 45-125 1000 n.r.

Ni 32 ±8.3 ζ 30 ±2.7 γ 30δ 16 10.7 8-28 30 2500

Sr 260 ±70 270 270 n.r. n.r. n.r. n.r. n.r.

Mn 210 ±55 370 400 n.r. n.r. n.r. n.r. n.r.

Mg 5800 ±1500 6300 4000 n.r. n.r. 8000-11000 n.r. n.r.

As 3.7 ±1.0 9.3 10 n.r. n.r. n.r. n.r. n.r.

Hg 0.26 ±0.069 0.3 n.a. 0.42 0.24 n.r. 0.8 200

Pb 32 ±8.3 27 ±3.3 18 54 17 10-60 120 10000

Sb n.d. n.d. n.d. n.r. n.r. n.r. n.r. n.r.

Zn n.m. 880 ±110 850 205 339 150-300 4000 n.r.

Co n.m. 3.9 4.6 n.r. n.r. n.r. n.r. n.r.

PAH n.m. 1.5 n.a. n.r. 0.47 n.r. 3 n.r.

DEHP n.m. 75 ±1 n.a. n.r. 61 n.r. 50 n.r.

LAS n.m. 130 n.a. n.r. 100 n.r. 1300 n.r.

NPs n.m. 9.6 ±0.35 n.a. n.r. 5.3 n.r. 10 n.r.

α Corresponding to 120 mg kg-1 P (limit 100, see 4b). β Corresponding to 110 mg kg-1 P (limit 100, see 4b). γCorresponding to 4700 mg kg -1 P (limit 2500, see 4b). δ Corresponding to 5000 mg kg-1 P (limit 2500, see 4b). εCorresponding to 78 ±53 mg kg -1 P (limit 100, see 4b). ζ Corresponding to 5300 ±2700 mg kg -1 P (limit 2500, see4b).

Alternative solutions to direct on land application might be: i) co-digestion of the

bioliquid with raw manure with further application on land of the digestate, ii)

incineration of the digestate and iii) post-composting of the digestate and disposal in

landfill. The first allows for recycling the nutrients though the decreased concentration

of metals and DEHP would only be an effect of dilution and the overall load unchanged.

The second and third prevent from loading agricultural soil with metals and DEHP;



28

The European Parliament and The Council, 2008. Directive 2008/98/EC of the European

Parliament and of the Council of 19 November 2008 on waste and repealing certain

Directives. The European Parliament and The Council.

Tonini, D., Astrup, T., 2012a. Life-cycle assessment of biomass-based energy systems: a case

study for Denmark, Appl. Energy 99, 234-246.

Tonini D., Hamelin L., Wenzel H., Astrup T., 2012 Bioenergy production from solid biomass: a

consequential LCA of 12 bioenergy scenarios. Env. Sc. Tec. DOI: 10.1021/es3024435.

Tonini D., Astrup T., 2012b. Life-cycle assessment of a waste refinery process for enzymatic

treatment of municipal solid waste. Waste Manage. 32, 165-76.

Trzcinski, A.P., Stuckey, D.C., 2012. Determination of the Hydrolysis Constant in the

Biochemical Methane Potential Test of Municipal Solid Waste, Environ. Eng. Sci. 29, 848-854.



V

Mechanical-biological treatment:performance and potentials. A LCA of 8

MBT plants including wastecharacterization

Montejo, C., Tonini, D., Marquez, C.M., Astrup, T.

Journal of Environmental Management, submitted



Mechanical-biological treatment: performance and potentials. A LCA of 8

MBT plants including wastecharacterization

Cristina Montejo 1*, Davide Tonini 2, María del Carmen Márquez 3, Thomas Astrup 4

1, 3 Department of Chemical Engineering, University of Salamanca, Plaza de los Caídos,1-5, 37008 Salamanca, Spain

2, 4 Department of Environmental Engineering, Technical University of Denmark, DTU,Miljoevej, Building 115, 2800 Kgs. Lyngby, Denmark

*Corresponding author : [email protected]



1

Abstract

In the endeavour of avoiding presence of biodegradable waste in landfills and increasing

recycling, mechanical-biological treatment plants have seen a significant increase in

number and capacity in the last two decades. The aim of these plants is separating andstabilizing the quickly biodegradable fraction of the waste as well as recovering

recyclables from mixed waste streams. In this study the environmental performance of

eight MBT-based waste management scenarios in Spain was assessed by means of life

cycle assessment. The focus was on the technical and environmental performance of the

MBT plants. These widely differed for type of biological treatment and recovery

efficiencies. The results indicated that the performance is strongly connected with

energy and materials recovery efficiency. Major potentials for improvement are

associated with optimizing materials recovery through increased automation of the

selection (particularly for metals and plastic) and with replacing direct composting by a

combination of anaerobic digestion and post-composting. Overall, up to ca. 177-188 kt

CO 2-eq. y -1 may be saved by optimizing the plants under assessment. For refuse derived

fuel (RDF) management, incineration induced higher greenhouse gas emissions

compared with the current management practice (landfilling) when the marginal

electricity source in the system was assumed as natural gas. However, RDF incineration

could benefit all the remaining environmental categories, particularly the groundwater

resource. The recommendation for upgrading and/or commissioning of future plants is

to optimize materials recovery through increased automation of the selection and to

prioritize biogas-electricity production from the organic fraction over direct composting.

The optimal strategy for RDF depends upon the environmental compartment to be

prioritized and on the type of marginal energy in the system.

Keywords : MBT, LCA, waste composition, biological treatment, material recovery,RDF



2

Glossary:

AC: Acidification

ETs: Ecotoxicity in soil

ETwc: Ecotoxicity in water (chronic)GHG: Greenhouse gas

GW: Global warming

HDPE: High density polyethylene

HTa: Human toxicity via air

HTs: Human toxicity via soil

HTw: Human toxicity via water

HRT: Hydraulic retention time

LCA: Life cycle assessment

LDPE: Low density polyethylene

MBP: Mechanical-biological pretreatment

MBS: Mechanical-biological stabilization

MBT: Mechanical-biological treatment

MBTP: Mechanical-biological treatment plant

MSW: Municipal solid waste

NE: Nutrient enrichment

OFMSW: Organic fraction of municipal solid waste

PET: Polyethylene terephthalate.

POF: Photochemical ozone formation

RDF: Refuse derived fuel

SGR: Spoiled groundwater resources

SOD: Stratospheric ozone depletion

VS: Volatile solids

WEEE: Waste electrical and electronic equipment

ww: wet waste



3

1. Introduction

In the last two decades about 180 MBT (mechanical-biological treatment) plants have

been installed in Europe (ECN, 2009) with the aim of avoiding the presence of

biodegradable waste in landfills, according to the European Directive 1999/31/CE(CEC, 1999). These plants combine mechanical separation of different fractions

contained in household waste with stabilization of organic matter by means of

biological processes such as anaerobic digestion or composting. According to Bilitewski

et al. (2011), two main types of MBT technology exist: A) mechanical-biological

pretreatment (MBP), where the organic fraction is separated and biologically stabilized

prior to landfilling and recyclables as well as RDF are recovered from the residual

coarse fraction, and B) mechanical-biological stabilization (MBS) or biodrying, which

first composts the waste for drying prior to extraction of a larger RDF fraction. MBP

aims at stabilizing the organic to minimize gas as well as leachate emissions in landfill

while MBS maximizes the RDF and materials recovery. Within this general

classification, multiple variations can be found and it can be stated that probably there

are no two identical plants.

The proliferation of these facilities was particularly remarkable in Spain where

the waste treatment capacity was increased by 5 million tonne by installation of 50 new

MBT plants (MMA, 2006). Particularly, in the region of Castilla y León (Spain), where

more than 1.2 million tonne of waste are generated annually, 11 MBT plants have been

built serving a population of 2.5 million of inhabitants within 94,223 km 2. It should be

noticed that selective source-segregation (separate collection of metals and plastic

containers along with glass and paper) corresponds to only 12% (as for 2009) of the

total collected waste (MMA, 2010), the remaining 88% being residual MSW. All the

residual MSW generated in this region is sent to MBT prior to landfilling.

Over the last ten years these plants have been in service, no specific assessments

on their environmental performances have been performed. To this respect life cycle

assessment (LCA) is a useful tool allowing for a holistic and systematic assessment of

both direct and indirect environmental impacts of a selected system. LCA includes

impact categories ranging from climate change (greenhouse gas (GHG) emissions) to

human health impacts associated with the release of toxic substances, and to



7

same approach as for the input rMSW (16 samples were collected; 19 waste material

fractions were sorted and weighted from the final sample of 250 kg). For the OFMSW,

due to the smaller size, (smaller than 90 cm), 16 samples of 50 kg were collected. Once

all these streams were analyzed, the amount of recyclable materials recovered in themechanical processes can be calculated by means of a global mass balance. Similarly,

transfer coefficients (i.e. distribution of the waste material fractions among the output

streams) for selected waste materials were calculated for each individual MBT plant.

2.4 Scope and functional unit definition

The functional unit of the LCA was the treatment of 1 tonne of wet rMSW (i.e. residual

MSW left-over after source-segregation at the household) with the purpose of

stabilizing the organic matter in order to achieve a final stabilized composted material

(BOE, 2005; Hogg et al., 2002). The geographical scope was Spain (Castilla y Leon).

The temporal scope can be proximate with ‘2012-2015’ as technologies efficiencies,

waste composition and amount, transport distances and marginals referred to current

knowledge and practices (in other words, no forecasting has been made). The

environmental impacts were assessed for a time horizon of 100 years, from the moment

when the rMSW was collected. The ‘zero burden’ approach was applied: all upstream

emissions associated with generating the waste were omitted from the LCA. The

chemical composition of each waste material fraction was assumed from Riber et al.

(2009). The boundary of the system was set at the point of rMSW collection; further

impacts and savings associated with downstream utilization of produced electricity (e.g.

from biogas or RDF combustion) and landfilling of residuals (e.g. rejects, RDF,

eventual incineration residuals, etc.) was accounted for by system expansion following

the principles of consequential LCA (Finnveden et al., 2009; ISO 2006a, ISO 2006b).

This implied that the products generated by the system (e.g. electricity and recycled

materials) substituted the relative marginal products in the market. In the particular case

of electricity, the marginal technology was assumed to be natural gas-fired power plants

based on MITYC (2011) and IDAE (2010). These indicate a trend on increasing the

energy provision by expanding the capacity of natural gas-based power production

while keeping coal and fuel oil steady. This assumption is also in accordance with other



9

One PE corresponds to the environmental load caused by one average EU-15 citizen in

one year covering all activities in life.

Table 1 . Normalization references in the LCA method EDIP97 (Stranddorf et al., 2005).Category Acronym

Physical basis

Normalizationreference EU-15

Unit

Global warming GW Global 8,700 kg CO 2-eq. person -1 y-1

Photochemical ozone formation POF Regional 25 kg C 2H4-eq. person -1 y-1

Stratospheric ozone depletion SOD Global 0.103 kg CFC-11-eq. person -1 y-1

Acidification AC Regional 74 kg SO 2-eq. person -1 y-1

Nutrient enrichment NE Regional 119 kg NO 3- -eq. person -1 y-1

Ecotoxicity in soil ETs Regional 964,000 m 3 soil person -1 y-1

Ecotoxicity in water chronic ETwc Regional 352,000 m3

water person-1

y-1

Human toxicity via soil HTs Regional 127 m 3 soil person -1 y-1

Human toxicity via water HTw Regional 50,000 m 3 water person -1 y-1

Human toxicity via air HTa Regional 60,900,000,000 m 3 air person -1 y-1

Spoiled groundwater resources SGR Regional 130 m 3 groundwater person -1 y-1

2.6 LCA scenarios

For each type of MBT ( I and II ), four corresponding waste management scenarios

(based on the eight MBT plants investigated) were assessed: MBTP I-1, MBTP I-2,MBTP I-3 and MBTP I-4 corresponding to type I and MBTP II-1, MBTP II-2, MBTP

II-3 and MBTP II-4 corresponding to type II . The LCA included two sets of scenarios:

the first set (a) consisted of the selected eight waste management scenarios (i.e. MBTP

I-1, MBTP I-2, MBTP I-3, MBTP I-4, MBTP II-1, MBTP II-2, MBTP II-3, MBTP II-4)

where the waste composition was specific for each individual plant (as investigated).

The second (b) consisted of the same eight waste management scenarios with the

difference that the waste input to the individual MBT plants was the same (average

waste composition for the region, see Table 2). The latter allows for comparing the

technical performance of the individual MBT plants (including further treatments)

disregarding the effects of different waste compositions. The system boundary is

exemplified in Figure 1 for the case of scenario MBTP I-3 (a). A total of 16 scenarios

were therefore assessed: 8 (scenarios, i.e. I-1 to II-4) x 2 waste compositions (i.e. a and

b) = 16.



10

In addition, the potentials for optimizing the environmental performance of the assessed

waste management scenarios were evaluated by analyzing the following changes in the

management system: i) RDF incineration instead of landfilling. ii) Optimization of the

biological treatment: this consisted on modelling the biological treatment in all thescenarios conformingly with the best performing digestion process (that is, MBTP I-2,

see Table 3). iii) Optimization of materials recovery: the potentials for GHG savings

associated with improved materials recovery at the MBT plants were quantified. It was

assumed that the MBT plants were upgraded with additional installation of the

following automatic selection units: optical separator for hard plastic (4.7 kWh tonne -1

ww), PET (1.5 kWh tonne -1 ww), soft plastic (assumed as for hard plastic), glass (glass

breaker and optical glass sorting, 20 kWh tonne -1 ww), aluminium (ECS, 0.88 kWh

tonne -1 ww) and ferrous metals (magnet, 2.4 kWh tonne -1 ww). Data were based on

Combs (2012). Paper and cardboard were included in the analysis (manual separation

was assumed). However, improved recovery of these may be limited by the

contamination with organic and the results have to be regarded only as upper potentials.

Figure 1 . Illustration of the LCA system boundary for the case of MBTP I-3 (a) ( MBT type I ).OFMSW: organic fraction of MSW (i.e. fine fraction screened by trommels). RDF: refuse

derived fuel. C sequestered refers to the carbon stored in landfill after the considered 100 yearshorizon. Values rounded to 2 significant digits.



13

2.7 Waste treatment technologies: LCI

2.7.1 Collection and transport

Collection was modelled by its diesel consumption per tonne of wet waste collected

(3.27 L diesel tonne-1

ww) according to Larsen et al. (2009a). Each recyclable fractionmechanically recovered in the MBT plant was baled, stored separately and sent to

recycling facilities. The fuel consumption for transportation to the recycling plants was

modelled based on Eisted et al. (2009). The fuel consumption ranged from 0.04 L diesel

km -1 tonne -1 RDF to 0.2 L diesel km -1 tonne -1 plastic. Distances were based on current

destination of the recovered products which are 200 km for glass, 250 km for paper, 300

km for metals, 700 km for transportation of both plastics and beverage cartons and 2 km

for transport of compost, rejects and RDF to the landfill.

2.7.2 MBT plants

Each individual MBT plant was modelled as a combination of mechanical separation

and biological treatment. The recovery efficiencies of the selected waste material

fractions were calculated based on the experimental data. The consumption of energy of

each machine and operation occurring at the plant under assessment was based on

Combs (2012). The average electricity consumption of the assessed plants for the

mechanical and manual operations was estimated to 15 kWh tonne -1 ww. The biological

treatments for both types of MBT plants were modelled according to Boldrin et al.

(2011) with respect to N 2O, CH 4 and NH 3 emissions. The purpose of the biological

treatment is the stabilization (aerobic or anaerobic or in combination) of the degradable

organic matter in order to achieve a final composted material having low methane

potential (typically < 20 Nm 3 tonne -1 ww) which allows sustainable landfilling with

decreased environmental impacts especially with respect to gas and leachate generation

(Stegmann, 2010; Cossu et al., 2003). The residual methane potential of the stabilized

organic matter was assumed 15 Nm 3 tonne -1 ww according to previous studies

(Manfredi et al., 2010a). The corresponding CH 4 emissions occurring in the landfill

were accounted for in the LCA.



15

MBT type IIBiological treatment in MBT type II was modelled as composting tunnels based on a

full-scale operating plant located in Italy (Boldrin et al., 2011). The degradation of each

material fraction was modelled as a percentage of the VS content in the incoming waste;

this corresponded to about 70% VS degradation for organic waste such as animal and

vegetable food waste, 60% for yard waste, 10-20% for paper and cardboard fractions

and 5% for beverage cartons and textiles. OFMSW and electricity consumption (about

20-25 kWh tonne -1 ww depending on the plant) were the main inputs to the composting

process whereas compost, rejects from refining processes and emissions of CH 4, N 2O

and NH 3 were the main outputs. According to Boldrin et al. (2011), the fugitive CH 4

emissions were set to 0.2% of the degraded carbon, N 2O emissions to 1.4% of thedegraded nitrogen. About 98.5% of the degraded nitrogen was in the form of NH 3 of

which 99% was assumed oxidized in biofilters (Boldrin et al. 2011) which are the

current air treatment system at the assessed plants. The compost and the rejects were

assumed landfilled according to current practices.

2.7.3 Recycling

Each individual waste material fraction recovered in the MBT plants was assigned a

specific recycling technology. The recycling processes were modelled by implementing

a combination of two parameters: technical substitution and market substitution. The

percentage of technical substitution represents the material loss in the recycling process

since losses occur during processing. The amount of recycled material is thus a

percentage of the input waste material. The percentage of market substitution is related

to the market acceptance of the recycled product. Recycled materials substitute similar

products made with virgin material avoiding impacts generated by the original

manufacturing; the market substitution is thus estimated as a percentage of the avoided

production.

Paper and aluminium recycling technologies were based on generic European

data (EDIP database). The material loss during the processes was set to 18% and 21%

respectively and market substitution was set to 100% for both materials according to

Schmidt and Strömberg (2006). Technologies for plastic recycling were based on data

from existing facilities in Denmark and Sweden. HDPE and LDPE substituted for



16

similar materials produced from virgin resources with an efficiency of 90% and the

material loss reached 10%. In the case of PET, recycled material avoided 100% virgin

production and technical substitution was set to 97%, i.e. 3% of material was lost during

the process. A glass recycling technology based on remelting of glass cullet tomanufacture new bottles was selected. Both technical and market substitutions were

assumed 100%. Ferrous materials recovered through magnets in MBT plants are

shredded and new sheets are made from the scraps; process and consumption were

based on a Swedish facility. Steel losses were set to 13% and the market substitution to

100%. The data of these processes were taken from the EDIP database. Additionally,

energy and GHG performances of these recycling processes can be found in: Merrild et

al. (2009) for paper and cardboard, Astrup et al. (2009a) and Bernstad et al. (2011) for

different plastic materials, Damgaard et al. (2009) for metals, Larsen et al. (2009b) for

glass.

With respect to beverage cartons, these were assumed composed of 74% paper,

22% polyethylene and 4% aluminium (Tetra Pak, 2004). The first step in recycling is to

divide the recovered cartons into their individual components. To this purpose, a

process similar to paper repulping is performed recovering the paper fraction by means

of water addition. However, process efficiency is lower than original repulping because

of the presence of polymer and aluminium layers. Beverage carton fiber has desirable

properties and can be made into folding boxboard, corrugation fluting and other

products. The cardboard market substitution was set to 90%. The mixed stream of

polyethylene and aluminium is subjected to pyrolysis processes recovering thus

aluminium powder which market substitution was assumed 100%. The heat resulting

from pyrolysis is used to dry the paper fibers. The overall material loss in the processes

was 22% corresponding to the plastic content of beverage cartons. The energy

consumption was set to 75 kWh tonne -1 of beverage cartons based on the life cycle

inventory of containers systems for wine (Franklin, 2006).

For example with respect to GHG emission savings the net (i.e. including the

environmental burdens of the recycling process itself) CO 2-eq. savings associated with

recycling 1 tonne of, respectively, aluminium, ferrous material, paper, hard and soft



17

plastic, plastic bottles, glass and beverage cartons equalled (assuming natural gas as

marginal) 8200, 5800, 720, 1200, 2400, 680, 260 and 1000 kg CO 2-eq. tonne -1 ww.

2.7.4 LandfillingStabilized organic material (compost), RDF and rejects from compost refinement were

landfilled. The landfill was modelled as a conventional landfill (Manfredi and

Christensen, 2009). This technology involved suitable collection systems of leachate

generated because of degradation, leachate treatment process, gas collection system,

flaring and oxidizing compost layer on the top of the landfilled waste. The assessed

100-year horizon was divided in four time periods to better represent the different

operational conditions of the landfill cells (Manfredi and Christensen, 2009). It was

assumed that 80% of the landfill gas generated was collected during the second

(duration: 8 years) and third period (duration: 35 years) whilst was not collected in the

first (duration: 2 years) and last (duration: 55 years). 100% of the collected gas was

assumed burnt in the flare as this is the current management scenario in the

geographical region assessed. Composition of leachate and landfill gas was defined

according to previous studies (Manfredi and Christensen, 2009).

2.7.5 RDF incineration

Although in the assessed reference scenarios RDF was landfilled (in accordance with

the current waste management system in the assessed geographical region), potential

changes in RDF management have been studied. This translated into dedicated

incineration of RDF in place of the current landfilling.

The incinerator was modelled based on the technology described in Tonini and

Astrup (2012a). The incinerator was assumed being a grate-fired incinerator equipped

with wet flue gas cleaning, selective non-catalytic reduction (SNCR) of NO x, Hg and

dioxin removal by activated carbon. The gross electricity efficiency of the incinerator

was assumed 30%, relative to the lower heating value (LHV) of the waste input,

representing state of the art incinerators combusting high-energy content materials and

provided with flue-gas condensation (DEA, 2012, International Solid Waste Association

(ISWA), 2006). Internal electricity consumption at the plant was 65 kWh tonne-1

ww



19

Table 4 . Recovery efficiencies (percent in wet basis, kg ww recovered kg -1 ww input): percentof individual waste materials (e.g. hard plastic, paper and cardboard, etc.) recovered at theassessed MBT plants. Values rounded to two significant digits.

Waste material fraction MBTPI-1

MBTPI-2

MBTPI-3

MBTPI-4

MBTPII-1

MBTPII-2

MBTPII-3

MBTPII-4

Organic matter 93 90 87 85 89 90 87 81Paper and cardboard 1.1 0.2 17 1.0 39 0.5 2.2 32

Hard plastic (HDPE) 0.7 24 34 2.7 8.7 1.5 4.6 36

Plastic bottles (PET) 11 28 33 4.9 43 40 48 24

Soft plastic (LDPE) 0.4 10 31 39 6.0 1.4 8.8 60

Glass 12 12 4.7 4.6 49 5.0 8.0 9.3

Ferrous metals 84 68 58 74 54 57 60 35

Aluminium metals 77 61 47 72 33 42 72 95

Beverage cartons 9.6 60 65 57 56 65 71 65

Table 5 . Biogenic carbon balance (kg C); rMSW: residual MSW input to the MBT plant; RDF:refuse derived fuel; CP: composted material (including rejects); Loss: biogenic C degraded (orfound in recycled materials); a: waste-specific results; b: waste-average results.

MBTPrMSW

Outputs of MBT Sequestration in landfillLoss

RDF CP RDF CP

a b a b a b a b a b a bI-1 183 183 71 77 39 37 53 58 9 8 121 117

I-2 176 183 68 87 37 33 50 63 9 8 117 112

I-3 190 183 75 68 36 37 54 50 8 8 128 125

I-4 169 183 72 84 33 34 50 60 8 8 111 116II-1 206 183 73 62 35 35 55 47 8 8 143 128

II-2 176 183 65 87 38 33 46 62 9 8 121 113

II-3 181 183 69 70 38 39 49 51 9 9 123 123

II-4 190 183 88 74 29 32 65 51 7 7 112 125

The LCA results for the eight scenarios are shown in Figure 2-3. These included the a)

‘waste-specific’ results (i.e. environmental impacts associated with the individual

scenarios provided the waste input is specific for each individual MBT plant) and b)‘waste-average’ results (where the waste input is the average for ‘Castilla y León’ for all

the scenarios). The latter allows for comparing the technical performance of the

individual MBT plants disregarding the effects associated with differences in the waste

compositions. The impacts of the principal processes involved have been detailed in the

charts. The processes were grouped into: i) Transportation (waste collection and

transport of recovered materials and rejects), ii) Mechanical treatment (mechanical

material recovery), iii) Biological processes (composting or combined process with



21

Figure 2 . Characterized environmental impacts on selected non-toxic categories: processcontributions. a: waste-specific results; b: waste-average results.



22

Figure 3 . Characterized environmental impacts on selected toxic categories: processcontributions. a: waste-specific results; b: waste-average results .



23

3.1 ‘Waste-specific’ results (current performance)

3.1.1 Non-toxic categories

On GW all the scenarios contributed with environmental savings. This was expected as

a zero-burden approach is assumed. However, the magnitude of the savings varieddepending on the scenario (which relates to the type of MBT plant, i.e. I or II ) and on

the type of marginal energy considered, that is, natural gas or coal. The choice of natural

gas determines that the energy recovery processes are less beneficial on GW compared

with choosing coal. This is also expected as coal has a significantly higher GHG

emission factor than natural gas (in this study: 1.1 vs. 0.49 kg CO 2-eq . kWh -1). This

finally implicates that, when natural gas is the marginal, most GHG savings are

provided by other processes than energy recovery’s, for example recycling or carbon

sequestration in the landfill. The overall GW savings ranged from -340 for MBTP II-1

(type II ) to -120 kg CO 2-eq. tonne -1 ww for MBTP II-2 ( type II ). There was no clear

evidence that scenarios with type I were better than II or vice versa. In fact, as opposed

to previous studies on similar subjects (e.g. Manfredi et al., 2011; Boldrin et al., 2011)

where anaerobic digestion of the organic matter was concluded to be favourable over

direct composting, in the present this was not the case. The primary reason for this was

the wide variation of energy and materials recovery efficiencies across the eight

scenarios. For example, MBTP I-1 and MBTP I-2 had high electricity recovery but

scarce materials recovery efficiencies. In addition, methane production was lower than

in the other MBT type II . Second, in the aforementioned studies coal was generally

assumed as marginal; this determined significantly higher benefits associated with the

energy recovery processes.

For all the scenarios the environmental savings were primarily associated with

(Figure 2a): (1) recycling, (2) landfilling (carbon sequestration) and (3) biological

processes (substitution of fossil fuel through energy produced during anaerobic

digestion). (1) Recycling determined significant savings in all scenarios; these were

largely dependent on the total amount of recovered materials (i.e. product of plant

separation efficiency and input material amount) particularly with respect to i) paper, ii)

plastic and iii) metals. The greatest savings were found for MBTP II-1, MBTP II-2 and

MBTP II-4 (-250, -240 and -230 kg CO 2-eq. tonne-1

ww, respectively) which also



24

showed the highest recovery efficiencies for these recyclables (Table 4). These results

are in agreement with previous studies focused on recycling of different materials

(Larsen et al., 2009b; Manfredi et al., 2011; Merrild et al., 2009; Merrild et al., 2012;).

(2) Landfilling of compost (that is, stabilized organic) and RDF resulted in significantGW benefits for all the scenarios (ranging between -200 (MBTP II-4) and -140 (MBTP

I-4) kg CO 2-eq. tonne -1 ww); these savings were completely associated with carbon

sequestration whereas minor impacts (about 15-25 kg CO 2-eq. tonne -1 ww) were a

consequence of residual methane emissions and energy consumption for the operations.

Between 84% and 91% of the total biogenic carbon sequestered was associated with the

RDF and between 9% and 16% with compost and rejects from the biological treatment

(see Table 5). These results are in agreement with the findings of previous studies (e.g.

Manfredi et al., 2009, Manfredi et al., 2010b, and Manfredi et al., 2011) where potential

savings associated with carbon sequestration were illustrated for different waste types.

(3) Biological processes resulted in GW savings only for MBTP I-2 when natural gas

was the marginal thanks to the significant electricity recovery. When coal was the

marginal also MBTP I-1 achieved net GHG savings. For all the scenarios, the GHG

emissions (CO 2-eq. tonne -1 ww) ranged from -7 to 48 in the case of natural gas and -48

to 75 kg in the case of coal. As shown by the results for MBTP I-3 and MBTP I-4, a

gross electricity efficiency of 13-14% was not sufficient to assure ‘GHG-neutrality’ to

the biological treatment itself (i.e. to compensate the impacts due to the emissions and

consumptions) neither in the case of natural gas nor coal as marginal. However, an

optimized use of the biogas-energy through, for example, maximization of the

electricity production and later use of the waste heat from the gas engine to heat the

digestion process may drastically increase the associated GHG savings. Additional

environmental impacts were caused by mechanical treatment to separate organic matter

and recyclables (ca. 8.5 kg CO 2-eq. tonne -1 ww due to energy consumption) and

transportation (between 17 and 40 kg CO 2-eq. tonne -1 ww depending on the amount of

recyclables transported).

The results for AC and NE followed a similar trend among the assessed

scenarios. On both categories recycling was the most important contributor to the

savings. All the other processes determined environmental impacts, primarily related to



25

(the order reflecting the relevance): NH 3 emissions from composting (biological

processes), NH 3 and PO 43- from landfilling (leachate) and NO x (transportation). On NE

NO x and NH 3 emissions exceeded the avoided N-emissions associated with recycling

determining an overall impact on this category in all the addressed scenarios. This wasnot the case for AC where only one scenario performed with a net impact (MBTP II-2)

due to scarce materials recovery efficiencies at the MBT. This could be avoided by, for

example, optimizing metals (particularly aluminium) recovery.

Regarding the remaining non-toxic categories, no remarkable differences among

the scenarios were found in the categories SOD and SGR (Table 6). Here the impact

was totally caused by landfilling (due to CFCs and ammonia emissions to groundwater

through leachate). On POF impacts were primarily caused by transportation (VOCs)

and to a minor extent by landfilling and biological processes (fugitive CH 4 emissions).

3.1.2 Toxic categories

As illustrated in Figure 3a, the results on the toxic categories ETwc, HTw and HTs

highlighted the benefits associated with recycling to which all the environmental

savings were connected across the different scenarios. As a consequence, the scenarios

with highly efficient recovery of recyclables at the MBT plant achieved the greatest

performances. Particularly, the recovery efficiency of aluminium determined the

performances on ETwc, HTs and, to a minor extent, also in HTw: the higher the overall

recovery at the MBT (see Table 4) the better the results; the impacts were mostly due to

transportation and landfilling as a consequence of metals emissions (through leachate)

and uncombusted hydrocarbons (transportation). On ETwc only the scenarios with

aluminium recovery efficiency at the MBT above 60% achieved environmental savings

(that is, MBTP I-1, MBTP I-2, MBTP I-4 and MBTP II-4); the impacts from transport

and landfilling were comparable for all scenarios. On HTw the performance was mainly

related to energy and aluminium recovery (and consequent avoided emissions from

fossil fuel combustion and virgin material production): the scenarios with direct

composting and scarce recovery efficiencies (that is, MBTP II-2, MBTP II-3 and MBTP

I-4) performed therefore worst. All the assessed scenarios performed with impacts on

HTa (Table 6); this was principally attributed to VOCs emissions from transportation.



26

The impacts on ETs (not shown) were negligible compared with the remaining

categories (Table 6).

3.2 ‘Waste-average’ results (current performance)Figure 2b-3b allows for comparing the scenarios disregarding the effects associated

with the differences in the waste composition thus gaining insight into the technical

performances of the MBT plants investigated. The results for GW highlighted that,

generally, the performance of the scenarios with MBT type I was better than type II

provided high electricity recovery (e.g. MBTP I-1 and MBTP I-2). It should be noted

how the overall magnitude of the GW performance was affected by the changes in the

waste composition. The paramount average variation of GHG savings between case a

and b (i.e. absolute average GHG variation of the eight scenarios) corresponded to about

22%. The largest variation was seen for MBTP II-2 (40% net increase of GHG savings)

as a consequence of the increased content of ferrous metals, paper and cardboard in the

input waste composition.

The results for AC and NE reflected the overall materials recovery (and

consequent recycling) efficiency: MBTP II-4 performed best thanks to higher recovery

for aluminium, plastic, cartons and paper compared with the remaining scenarios.

However, as opposed to AC, on NE all the scenarios contributed with environmental

impacts as a consequence of higher N-emissions from composting and landfills in

comparison with the avoided N-emissions connected to recycling. Yet, efficient material

recovery at MBT has the potential to mitigate the overall NE impact (e.g. see MBTP II-

4). The impacts on the remaining non-toxic categories (POF, SOD and SGR) were

primarily due to landfilling and transportation as earlier described; the magnitude was

comparable for all the scenarios (Table 6).

On ETwc and HTs the results were mainly affected by the amount of recovered

aluminium (which determines the most significant savings on these categories as earlier

described): the scenarios having the lowest recovery efficiency at the MBT (MBTP I-3,

MBTP II-1 and MBTP II-2, see Table 4) were the worst. It can be noticed that the

overall performance of MBTP I-1 and MBTP II-4 decreased in (b) compared with (a) as

a result of the diminished share of aluminium in the average ‘Castilla y León’ waste



27

composition compared with the specific waste input to these plants (Table 2). In HTw

the major savings were associated with recycling of paper and aluminium and also, to a

minor extent (when natural gas is marginal) energy recovery, as earlier described. The

combination of no-energy as well as low materials recovery determined net impacts inthe case of MBTP II-2. This also applied to MBTP I-4 and to all scenarios with MBT

type II when coal was the marginal. All the addressed scenarios contributed with

comparable impacts (see Table 6) on HTa; these were primarily attributed to

transportation as earlier described.



2 8

T a

b l e 6

. N o r m a l i z e d

L C A r e s u l t

( m P E t o n n e - 1 w w

) f o r a l l t h e a s s e s s e d n o n - t o x i c a n d

t o x i c e n v i r o n m e n t a l c a t e g o r i e s .

V a l u e s r o u n d e d

t o t w o s i g n i f i c a n t d i g i t s .

N G / C o a l : n a t u r a l g a s / c o a l a s m a r g i n a l f u e l f o r e l e c t r i c i t y p r o d u c t i o n .

G W

A C

N E

P O F

S O D

S G R

E T w c

H T a

H T s

H T w

a

b

a

b

a

b

a

b

a

b

a

b

a

b

a

b

a

b

a

b

M B T P I - 1

N G

- 2 7

- 3 3

- 1 . 6

- 1 . 7

4 . 6

4 . 7

1 . 2

0 . 9

2 . 3

1 . 2

3 , 0 0 0

3 , 1 0 0

- 5 7

- 2 3

2 . 0

2 . 1

- 2 0

- 1 2

- 1 . 7

- 1 . 6

C o a l

- 3 3

- 3 5

- 1 . 5

- 0 . 9

5 . 1

5 . 1

0 . 7

0 . 7

1 . 4

1 . 3

3 , 0 0 0

3 , 1 0 0

- 5 8

- 2 2

2 . 0

2 . 2

- 2 1

- 1 0

- 2 . 3

- 1 . 8

M B T P I - 2

N G

- 2 4

- 3 4

- 3 . 4

- 4 . 0

4 . 5

3 . 9

1 . 7

1 . 2

2 . 4

1 . 5

3 , 0 0 0

3 , 1 0 0

- 1 6

- 1 6

2 . 5

2 . 5

- 8 . 6

- 1 0

- 0 . 8 8

- 1 . 3

C o a l

- 3 0

- 3 6

- 3 . 1

- 2 . 8

4 . 5

4 . 5

1 . 1

1 . 1

1 . 4

1 . 7

3 , 0 0 0

3 , 1 0 0

- 1 7

- 1 5

2 . 4

2 . 5

- 9 . 0

- 8 . 2

- 1 . 5

- 1 . 6

M B T P I - 3

N G

- 3 6

- 2 7

- 5 . 6

- 5 . 4

3 . 3

3 . 6

1 . 7

1 . 6

2 . 6

1 . 2

3 , 1 0 0

3 , 0 0 0

0 . 6

- 0 . 6

3 . 0

2 . 9

- 5 . 7

- 7 . 4

- 2 . 2

- 0 . 7 6

C o a l

- 3 0

- 2 4

- 5 . 6

- 4 . 9

4 . 6

4 . 3

2 . 1

1 . 7

1 . 6

1 . 4

3 , 1 0 0

3 , 0 0 0

1 2

2 . 1

4 . 2

3 . 1

- 6 . 1

- 5 . 8

- 0 . 7 4

- 0 . 1 2

M B T P I - 4

N G

- 2 3

- 2 7

- 3 . 1

- 1 . 9

4 . 5

4 . 8

2 . 0

1 . 5

2 . 5

1 . 6

3 , 1 0 0

3 , 1 0 0

- 2 4

- 1 9

2 . 9

2 . 7

- 1 2

- 1 1

- 0 . 3 1

- 0 . 2 9

C o a l

- 2 2

- 2 4

- 3 . 5

- 1 . 8

4 . 7

5 . 3

1 . 7

1 . 6

1 . 6

1 . 8

3 , 1 0 0

3 , 1 0 0

- 2 2

- 1 7

3 . 1

2 . 9

- 1 2

- 9 . 2

0 . 3 2

0 . 3 4

M B T P I I - 1

N G

- 3 9

- 2 3

- 4 . 3

- 4 . 5

2 . 1

3 . 1

1 . 3

1 . 3

1 . 9

1 . 0

2 , 4 0 0

2 , 3 0 0

7 . 8

5 . 7

2 . 3

2 . 6

- 3 . 3

- 4 . 8

- 3 . 1

- 0 . 5 2

C o a l

- 2 1

- 1 7

- 6 . 2

- 4 . 6

2 . 6

3 . 3

1 . 7

1 . 4

1 . 2

1 . 1

2 , 4 0 0

2 , 3 0 0

1 1

7 . 1

2 . 8

2 . 7

- 4 . 8

- 4 . 6

0 . 1 9

0 . 3 9

M B T P I I - 2

N G

- 1 4

- 2 5

1 . 2

- 0 . 3

5 . 5

5 . 1

1 . 6

1 . 3

2 . 1

1 . 6

2 , 3 0 0

2 , 6 0 0

4 . 6

- 5 . 9

2 . 5

2 . 5

- 0 . 2

- 4 . 6

0 . 6 5

0 . 3 1

C o a l

- 1 2

- 2 0

0 . 7

- 0 . 2

5 . 5

5 . 2

1 . 3

1 . 4

1 . 4

1 . 7

2 , 3 0 0

2 , 6 0 0

5 . 6

- 5 . 0

2 . 6

2 . 6

- 0 . 9

- 4 . 3

1 . 2

0 . 8 6

M B T P I I - 3

N G

- 2 4

- 2 3

- 3 . 5

- 2 . 9

3 . 7

4 . 2

1 . 4

1 . 1

2 . 3

1 . 2

2 , 6 0 0

2 , 6 0 0

5 . 9

- 2 3

2 . 7

2 . 6

- 1 . 7

- 1 1

- 0 . 2 3

- 0 . 2 1

C o a l

- 2 2

- 1 9

- 4 . 3

- 3 . 1

3 . 7

4 . 3

1 . 2

1 . 2

1 . 4

1 . 4

2 , 6 0 0

2 , 6 0 0

7 . 1

- 2 2

2 . 8

2 . 6

- 2 . 4

- 1 1

0 . 4 6

0 . 4 5

M B T P I I - 4

N G

- 3 8

- 2 5

- 9 . 2

- 7 . 9

1 . 4

2 . 1

1 . 9

2 . 0

2 . 2

1 . 4

2 , 6 0 0

2 , 4 0 0

- 6 9

- 2 2

3 . 1

3 . 3

- 2 8

- 1 5

- 2 . 4

- 0 . 4 1

C o a l

- 2 4

- 1 8

- 1 1

- 8 . 6

1 . 7

2 . 2

2 . 2

2 . 1

1 . 6

1 . 5

2 , 6 0 0

2 , 4 0 0

- 6 5

- 2 0

3 . 5

3 . 5

- 3 0

- 1 6

0 . 0 6

0 . 6 9



29

3.3 Potentials for optimizing the environmental performance

Figure 4-5 shows the environmental consequences in terms of benefits and/or impacts

associated with potentially improved waste management scenarios (i, ii and iii). The

bars in Figure 4 indicate the net difference in the environmental impact or saving between the improved and the baseline scenario that is set to zero and used as reference

(i.e. ∆saving or ∆impact = improved value – baseline value): a bar pointing towards

negative values indicates that the improved scenario achieved increased environmental

savings compared with the baseline (and viceversa). The waste-specific scenarios were

used as baseline. Figure 4(i) shows the variation in the environmental performance in

the case that RDF was incinerated instead of landfilled. Figure 4(ii) shows the variation

when the organic matter is anaerobically digested with optimized energy recovery (see

section 2.6) instead of ‘directly’ composted.

Figure 5 shows the paramount average GHG emissions (i.e. average of the eight

scenarios) associated with 100% material recovery at the MBT or, alternatively, 100%

material incineration, relative to the baseline LCA results. These were set to zero and

used as reference. Thus, any value below zero represents a saving compared with the

baseline (and viceversa). Table 7 reports the detailed results for the individual scenarios.

This analysis was performed to illustrate the potentials for GHG savings associated with

the management of the individual waste materials.

The results (Figure 4-5 and Table 7) showed that from a GHG perspective

significant environmental improvements may be achieved by a combination of the

following (the order reflecting the relevance): 1) increased materials recovery (primarily

metals and plastic) and 2) optimized biological treatment of OFMSW. Incineration of

RDF (instead of landfilling) was beneficial only if coal was the marginal. Overall,

optimization of metals and plastic recovery should be prioritized. This would enhance

the environmental performance on all environmental categories (exemplified in Figure 5

for GW). The associated savings varied across the scenarios depending upon the waste

composition. In the case of paper and cardboard, additional savings from improved

recovery are dramatically dependent upon the assumptions regarding C sequestration

and therefore the considered time horizon. If C sequestered was not accounted, the



30

additional savings from paper recycling compared with landfilling would be much more

significant (Figure 5 and Table 7).

Optimizing the biological treatment also induced additional GW savings

(between -8 and -93 kg CO 2-eq. tonne-1

ww). Note that biogas may also provideadditional benefits in relation to storability and flexibility of use in the perspective of

future energy systems with increased penetration of wind and other fluctuating

renewables such as photovoltaic and tides (among the others: Lund, 2007; Mathiesen et

al., 2011a; Mathiesen et al., 2011b; Tonini and Astrup, 2012b).

As aforementioned RDF incineration decreased the GW performances compared

with landfilling when C sequestration was accounted for along with natural gas being

the marginal. The reason for this was that the CO 2 savings from avoided natural gas

combustion were largely compensated by waste-specific CO 2 emissions (from plastic).

As opposite to this, the GHG performance was instead improved compared with the

baseline (landfilling) when coal was the marginal (Figure 4). This indicates that RDF

should be used to substitute for fossil fuels with higher emission factor than that of coal

(see also 3.3.1). Additionally, RDF incineration strongly mitigated the impact potential

on all the remaining categories, particularly on SGR. The latter was on average reduced

down to ca. 1/3 of the current impact potential. For example, for the case of scenario

MBTP II-2 the overall potential for GHG emission savings may equal ca. -240 kg CO 2-

eq. tonne -1 ww (-190 from recovery/recycling of paper and cardboard, plastic, and

metals and -50 from optimized biological treatment). RDF Incineration would induce a

net impact of ca. 140 kg CO 2-eq. tonne -1 ww with natural gas as marginal, and,

viceversa, a net saving of ca. -90 kg CO 2-eq. tonne -1 ww with coal as marginal.



31

Figure 4. Environmental consequences (selected non-toxic categories) associated with the potentially improved waste management scenarios (i and ii). The error bars indicate the net

difference in the environmental impact or saving between the improved and the baselinescenario (i.e. ∆saving or ∆impact = improved value – baseline value).



32

Figure 4 (cont.). Environmental consequences (selected toxic categories) associated with the potentially improved waste management scenarios (i and ii). The error bars indicate the netdifference in the environmental impact or saving between the improved and the baselinescenario (i.e. ∆saving or ∆impact = improved va lue – baseline value).



34

3.3.1 Energy from RDF: effects of energy efficiency, plastic recovery and fuel

substituted on GW

As aforementioned RDF incineration resulted in worse GW performance compared with

the baseline (landfilling) when natural gas was the marginal. This may change if: 1) theelectricity efficiency of the combustion process increases, 2) sorting and recovery of

plastic at the MBT is applied prior to RDF incineration, and 3) the substituted fuel is

simply different than natural gas. The first (1) was assessed by identifying the (net)

electricity efficiency ( η power plant ) that should be achieved at the dedicated RDF

incinerator (or a generic power plant) in order to equal the same GHG savings of the

baseline (where plastic along with the remaining RDF is landfilled and carbon from

paper and organic is sequestered). The calculation was performed according to Equation

1. Similarly, for the second (2), the plastic recovery efficiency ( ηrec) that should be

achieved at the individual MBT plants to equal the GHG performance of the baseline

was evaluated (only sorting of PET, soft and hard plastic were considered). The

calculation was performed according to Equation 2. The third (3) was exemplified in the

case of coal as marginal (section 3.3). However, Eq. 3 allows recalculating the specific

CO 2 emission factor that the ‘substituted fuel’ (EF fuel) should have in the individual

MBT scenarios to equal the GHG performance of the baseline. As highlighted in Table

8-9, a net electricity efficiency of 49-53% should be achieved to equal the GHG

performance of the baseline. Alternatively, a recovery efficiency at the MBTs greater

than 85-100% should be achieved for the considered plastic fractions. The CO 2

emission factor of the ‘fuel substituted’ (not shown) varied between 67 and 72 kg CO 2-

eq. MJ -1 depending on the scenario. This corresponds to the range of light oil fuels.

Thus, the substitution of any fuel having emission factor above this range would induce

net GHG savings compared with the baseline (RDF landfilling). This might be the case,

for example, of substitution of heavy fuel oil in cement kilns (emission factor ca. 80 kg

CO 2-eq. MJ -1).



35

12/446.3/ ⋅⋅+⋅⋅⋅−= RDF C EF RDF LHV GHG foss NG plant power RDF baseline η

Eq. 1

Where:

GHG baseline : GHG savings achieved in the baseline (landfilling RDF) (kg CO 2)

RDF: amount of RDF combusted (kg ww)

LHV RDF : LHV of the RDF (MJ/kg -1 ww)

η power plant : electricity efficiency of the power plant (%) (unknown)

EF NG : assumed GHG emission factor for natural gas (4.9 kg CO 2 kWh -1)

C foss: C fossil content in the RDF (kg C kg -1)

Notice that the only unknown is η power plant. The minus is used to maintain consistency

with the results discussion text where environmental savings are reported as negative

values. The C foss content of the material fractions constituting the RDF was assumed

according to Riber et al. (2009) as earlier reported (section 2.4). The terms 3.6 and

44/12 are conversion factors (MJ to kWh and C to CO 2).

)(

12/44*6.3/**

recycl HP recycl SP recycl PET rec

foss NGinc RDF baseline

Pot Pot Pot

RDF C EF RDF LHV GHG

++⋅−

⋅⋅+⋅⋅⋅−=

η

η

Eq. 2

Where:


RDF*: amount of RDF recalculated without the plastic sorted (kg ww)

LHV RDF* : LHV of the RDF* recalculated without the plastic sorted (MJ/kg -1 ww)

ηinc : electricity efficiency of the incinerator (30%, see 2.7.5)

EF NG : assumed GHG emission factor for natural gas (4.9 kg CO 2 kWh-1

)C foss: C fossil content in the RDF* (kg C kg -1)

ηrec: recovery efficiency for the plastic material fractions (%) (unknown)

Pot PET recycl : potential GHG saving associated with 100% recovery of PET (kg CO 2

tonne -1 ww)

Pot SP recycl : potential GHG saving associated with 100% recovery of soft plastic (kg CO 2

tonne -1 ww)



36

Pot HP recycl : potential GHG saving associated with 100% recovery of hard plastic (kg

CO 2 tonne -1 ww)

Pot PET recycl , Pot SP recycl , Pot HP recycl can be found in Table 7. The only unknown in Eq. 2 is

ηrec. The minus is used to maintain consistency with the text where savings are asnegative values and impacts as positive.

12/44⋅⋅+⋅⋅−= RDF C EF RDF LHV GHG foss fuel RDF baseline

Eq. 3


RDF: amount of RDF combusted (kg ww)

LHV RDF : LHV of the RDF (MJ/kg -1 ww)

EF fuel: emission factor of the fuel substituted (kg CO 2 MJ -1) (unknown)

C foss: C fossil content in the RDF (kg C kg -1)



3 7

T a

b l e 8

. E n e r g y f r o m R D F : e f f e c t s o f

i n c r e a s e d e l e c t r i c i t y e f f i c i e n c y o f

t h e p o w e r p l a n t .

L H V

R D F :

L H V

o f t h e

R D F ; η

e l : e l e c t r i c i t y e f f i c i e n c y ; E F N G : C

O 2

e m i s s i o n f a c t o r o f n a t u r a l g a s m a r g i n a l .

R e s u l t s a r e

b a s e d o n

E q . 1 .

M B T P

G H G

b a s e l i n e ( k g C O

2 )

R D F ( t o n n e )

L H V

R D F

( M J k g - 1 )

η e l

C f o s s · R

D F · 4 4 / 1 2 ( k g C O

2 )

L H V

R D F · R

D F · η

e l / 3 . 6 · E F N

G ( k g C O

2 )

I - 1

- 1 6 0

0 . 3 3

1 7 . 7

5 2 %

2 5 0

- 4 1 0

I - 2

- 1 5 0

0 . 3 1

1 5 . 3

5 1 %

1 8 0

- 3 3 0

I - 3

- 1 6 0

0 . 3 5

1 5

5 0 %

2 0 0

- 3 6 0

I - 4

- 1 4 0

0 . 3 7

1 5 . 2

5 0 %

2 4 0

- 3 8 0

I I - 1

- 1 7 0

0 . 3 4

1 6

5 3 %

2 1 0

- 3 8 0

I I - 2

- 1 3 0

0 . 3 3

1 3 . 6

4 9 %

1 7 0

- 3 0 0

I I - 3

- 1 4 0

0 . 3 2

1 4 . 6

4 9 %

1 8 0

- 3 1 0

I I - 4

- 1 9 0

0 . 3 9

1 3 . 5

5 0 %

1 7 0

- 3 6 0

T a b l e 9 . E n e r g y f r o m R D F : e f f e c t s o f p l a s t i c s o r t i n g a n d r e c o v e r y a t t h e

M B T p r i o r t o

R D F i n c i n e r a t i o n . A v a l u e

> 1 0 0 % m e a n s

t h a t t h e s e p a r a t i o n o f

r e c y c l a b l e p l a s t i c

i s n o t e n o u g h

t o e q u a l t h e

G H G s a v i n g s o f t h e

b a s e l i n e a n d a d

d i t i o n a l r e c o v e r y o f n o n - r e c y c l a b

l e p

l a s t i c w o u

l d b e n e e d e d .

η r e c : e f f i c i e n c y

o f

p l a s t i c s e p a r a t i o n .

L H V

R D F * : L H

V o f t h e

R D F * . R e s u l t s a r e

b a s e d o n

E q . 2 .

M B T P

G H G

b a s e l i n e ( k g C O

2 )

R D F ( t o n n e )

L H V

R D F *

( M J k g - 1 )

η r e c

( % )

G H G R D F * i n c i n e r a t i o n

G H G p l a s t i c r e c y c l i n g

I - 1

- 1 6 0

0 . 2 7

1 3 . 7

> 1 0 0 % ( > 6 7 % ) b

- 5 2

- 9 6

I - 2

- 1 5 0

0 . 2 7

1 3

9 5 % ( 7 8 % ) b

- 4 1

- 1 1 0

I - 3

- 1 6 0

0 . 3

1 2 . 9

> 1 0 0 % ( > 6 7 % ) b

- 4 6

- 1 0 0

I - 4

- 1 4 0

0 . 3 3

1 2 . 9

9 0 % ( 6 5 % ) b

- 2 4

- 1 2 0

I I - 1

- 1 7 0

0 . 2 9

1 3

> 1 0 0 % ( > 7 0 % ) b

- 4 2

- 8 5

I I - 2

- 1 3 0

0 . 2 9

1 0 . 9

9 5 % ( 6 7 % ) b

- 5 0

- 8 0

I I - 3

- 1 4 0

0 . 2 9

1 2 . 4

8 5 % ( 6 5 % ) b

- 4 6

- 1 1 0

I I - 4

- 1 9 0

0 . 3 7

1 2

> 1 0 0 % ( > 7 2 % ) b

- 6 2

- 1 1 0

b O v e r a l l e f f i c i e n c y o f p l a s t i c r e c o v e r y e f f i c i e n c y c a l c u l a t e d o n

t h e t o t a l p l a s t i c i n p u t ( r e c y c l a b l e p l u s n o t r e c y c l a b l e ) .



38

3.4 Perspectives

The results of this research suggest that the assessed MBT plants, even if recently

commissioned, aim at a safe disposal (e.g. stabilization of the OFMSW prior to

landfilling) rather than at maximizing energy and materials recovery. In this perspective, large potentials exist to optimize the environmental performance, thereby

changing the perception on MSW from ‘waste to be disposed’ to ‘resource and energy

carrier’. Table 10 presents an overview of the total annual GHG savings that could be

achieved in each individual plant by optimizing biogas-energy and materials recovery

(RDF is assumed landfilled). A total is also estimated. Overall, the estimated potential

for GHG emission savings equalled ca. 177111-187858 tonne CO 2-eq. y- 1 depending on

the assumption for the marginal electricity source.

The fact that the biggest savings are associated with recovery/recycling along

with the scarce GHG performance of RDF incineration (under the assumptions made

about marginal electricity source and carbon sequestration) highlight that optimization

of materials recovery is crucial, if the focus is mitigating GHG emissions. Future studies

should thus focus on evaluating alternative options to maximize recycling: these should

include source-segregation strategies to be integrated with optimized mechanical-

biological treatment. In addition, optimal strategies for energy recovery from RDF

should also be evaluated in more detail, possibly involving incineration with recovery of

the heat (for example for industry or for district heating/cooling), co-firing in dedicated

large scale power plants or in cement kilns, to maximize the total energy recovery.

Table 10 . Overall potentials for GHG savings. NG/CO: natural gas/coal as marginal forelectricity production.

MBTPMaterialrecovery

Biologicaltreatment

rMSWtreated

Potential GHG savings(tonne CO 2-eq. y -1)

NG CO NG CO tonne y -1 CO NG

I-1 -140 -74 -8 -46 309996 -47208.9 -37476.6

I-2 -140 -78 0 0 85764 -11816.3 -6656.4

I-3 -220 -150 -26 -74 83698 -20937.0 -18571.6

I-4 -160 -92 -33 -100 275432 -52638.7 -52814.9

II-1 -180 -110 -93 -120 27869 -7649.4 -6389.3

II-2 -120 -67 -54 -150 105731 -18579.7 -23162.5

II-3 -210 -140 -45 -130 46873 -12122.3 -12716.4

II-4 -140 -89 -38 -120 94802 -16905.5 -19323.1TOTAL 1030165 -187858 -177111



39

4. Conclusion

The current environmental performance of eight waste management scenarios based on

MBT in Spain was evaluated by means of LCA modeling. Potential benefits associated

with optimization of the management were assessed. From a GHG perspective theresults revealed that the environmental performance of the current waste management is

primarily connected with materials recovery, carbon sequestration in landfill and energy

recovery through anaerobic digestion of the organic matter. The waste composition

principally affected the magnitude of the benefits associated with recycling. These

varied widely across the assessed scenarios depending on both waste composition and

materials recovery efficiencies at the MBT. On the other categories the environmental

profile was primarily dictated by the amount of recovered materials.

The technical performance of the MBT plants was compared using an average

waste composition as input to all the scenarios. This highlighted the correlation between

materials/energy recovery efficiencies at the MBT and associated environmental

savings in all the addressed environmental categories. In particular, MBT plants with

efficient electricity recovery from biogas performed better. High recovery efficiencies

for paper, plastic and metals determined significant environmental savings on the non-

toxic categories whereas high recovery efficiencies for aluminium (including from

cartons) induced considerable benefits on the toxic.

Significant environmental improvements may be achieved by optimizing

materials recovery and replacing direct composting with a combination of anaerobic

digestion and post-composting. For RDF, possible GHG savings associated with energy

recovery from incineration are dependent upon assumptions regarding carbon

sequestration and surrounding energy system (marginal electricity source). However,

RDF incineration may significantly mitigate the impact potential in all the remaining

impact categories. The recommendation for upgrading and/or commissioning of future

plants is therefore to optimize materials recovery through increased automation of the

selection and to prioritize biogas-electricity production over direct composting of the

OFMSW with further re-use of the waste heat within the digestion process. The optimal

strategy for RDF depends upon the environmental compartment to be prioritized and on



40

the type of marginal energy in the system. To this respect, further investigations are

needed.

5. Acknowledgements

The authors gratefully acknowledge the financial support from the Junta de Castilla yLeón (Spain).

6. References

Abeliotis, K., Kalogeropoulos, A., Lasaridi, K., 2012. L ife Cycle Assessment of the MBT plantin Ano Liossia, Athens, Greece. Waste Manage. 32, 213-219.

ADEME, 1998. Agence de l'Environnement et de la Maîtrise de l'Energie. MODECOM TM:Method for characterization of domestic waste. ADEME. Paris, pp. 61.

Astrup, T., Fruergaard, T, Christensen, T.H., 2009a. Recycling of plastics: accounting ofgreenhouse gases and global warming contributions. Waste Manage Res. 27, 763-772.

Astrup, T., Møller, J., Fruergaard, T., 2009b. Incineration and co-combustion of waste:accounting of greenhouse gases and global warming contributions. Waste Manage Res. 27,789–799.

Bernstad, A., Jansen, la J., Aspegren, H. 2011. Life cycle assessment of a household solid waste

source separation programme: a Swedish case study. Waste Manage. Res. 29(10), 1027-1042.

Bilitewski, B., Oros, C., Christensen T.H. 2011. Mechanical Biological Treatment, inChristensen T.H. (Ed), Solid Waste Technology and Management. Willey & Sons, London.

Birgisdottir, H., Bhander, G., Hauschild, M.Z., Christensen, T.H., 2007. Life cycle assessmentof disposal of residues from municipal solid waste incineration: recycling of bottom ash in roadconstruction or landfilling in Denmark evaluated in the ROAD-RES model. Waste Manage. 27,S75–S84.

BOE, 2005. RD 824/2005 on fertilizer products. Boletín Oficial del Estado de 19 de julio de2005, 171, 25592-25669. Available on web: <www.boe.es/boe/dias/2005/07/19/pdfs/A25592-25669.pdf>

Boldrin, A., Neidel T.L., Damgaard, A., Bhander, G.S., Møller, J. and Christensen, T.H., 2011.Modelling of environmental impacts form biological treatment of organic municipal waste inEASEWASTE. Waste Manage. 31, 619-630.

Börjesson, P. and Berglund, M. 2006. Environmental systems analysis of biogas systems - PartI: Fuel-cycle emissions. Biomass Bioenerg. 30, 469-485.

http://redirect.cvt.dk/?url=http%3A%2F%2Fdtu-ftc.cvt.dk%2Fcgi-bin%2Ffulltext%2Felsevier%3Fpi%3D%2F0956%2F053x%2F00320001%2F11003801.pdf%26key%3D313044924%26rfr_id%3Dinfo%3Asid%2Fdlib.dtu.dk%3ADTUDigitalLibrary






41

Bovea, M.D., Ibáñez-Forés, V., Gallardo, A, Colomer-Mendoza, F.J., 2010. Environmentalassessment of alternative municipal solid waste management strategies. A Spanish case study.Waste Manage. 30, 2383-2395.

CEC, 1999. Council Directive 1999/31/EC on the landfill of waste. Official Journal of the

European Communities L 182/1. Available at:http://www.central2013.eu/fileadmin/user_upload/Downloads/Document_Centre/OP_Resources/Landfill_Directive_1999_31_EC.pdf (accessed 12.12).

Christensen, T.H., Gentil, E., Boldrin, A., Larsen, A.W., Weidema, B.P, Hauschild M., 2009. C balance, carbon dioxide emissions and global warming potentials in LCA-modelling of wastemanagement systems. Waste Manage. Res. 27, 696-706.

Combs, A.R., 2012. Life Cycle Analysis of Recycling Facilities in a Carbon Constrained World.

MS thesis, NC State Dept of Civil, Construction, and Environmental Engineering, NorthCarolina, USA.

Cossu, R., Raga, R., Rossetti, D., 2003. The PAF model: an integrated approach for land llsustainability. Waste Manage. 23, 37-44.

Damgaard, A., Larsen, A.W, Christensen, T.H., 2009. Recycling of metals: accounting ofgreenhouse gases and global warming contributions. Waste Manage. Res. 27, 773-780.

Danish Energy Agency (DEA), 2012. Technology data for energy plants - Generation of

Electricity and District Heating, Energy Storage and Energy Carrier Generation and Conversion.Danish Energy agency, Copenhagen, Denmark. Available at:http://www.ens.dk/Documents/Netboghandel%20-%20publikationer/2012/Teknologikatalog_2012.pdf (accessed 12.12).

ECN, 2009. Personal Communication with Josef Barth, Managing Director, European Compost Network (ECN).

Eisted, R., Larsen, A.W., Christensen, T.H., 2009. Collection, transfer and transport of waste:accounting of greenhouse gases and global warming contribution. Waste Manage. Res. 27, 738– 745.

Finnveden, G., Hauschild, M.Z., Ekvall, T., Guine, J., Heijungs, R., Hellweg, S., Koehler,A.,Pennington, D., Suh, S., 2009. Recent developments in Life Cycle Assessment. J. Environ.Manage. 91, 1–21.

Franklin Associates, 2006. Life cycle inventory of container systems for wine. Final report prepared for Tetra Pak Inc. by Franklin Associates, a division of ERG Prairie village, KS.

Fruergaard, T., Astrup, T., 2011. Optimal utilization of waste-to-energy in an LCA perspective.Waste Manage. 31, 572–582.



42

Fruergaard, T., Hyks, J., Astrup, T., 2010. Life-cycle assessment of selected managementoptions for air pollution control residues from waste incineration. Sci. Total Environ. 408,4672–4680.

Hamelin, L.; Wesnaes, M.; Wenzel, H.; Petersen, B.M. (2011). Environmental Consequences of

Future Biogas Technologies Based on Separated Slurry. Environ. Sci. Technol. 45, 5869-5877.

Hogg, D., Barth, J., Favoino, E., Centemero, M., Caimi, V., Amlinger, F., Devliegher, W.,Brinton, W., Antler, S., 2002. Comparison of compost standards within the EU, Northamericaand Australasia” Main report. The Waste and Resources Action Programme. Available at:http://www.compostingvermont.org/pdf/WRAP_Comparison_of_Compost_Standards_2002.pdf(accessed 12.12).

IDAE (2010). Spain’s national renewable energy action plan 2011-2020. Available at:http://pvtrin.eu/assets/media/PDF/EU_POLICIES/National%20Renewable%20Energy%20Acti

on%20Plan/202.pdf or at http://ec.europa.eu/energy/renewables/action_plan_en.htm (accessed12.12).

ISO 2006a. Environmental Management-Life Cycle Assessment-Principles and Framework.2nd ed. 2006; ISO 14040.

ISO 2006b. Environmental Management-Life Cycle Assessment-Requirements and Guidelines.1st ed. 2006; ISO 14040.

International Solid Waste Association (ISWA). Energy from Waste. State of the Art Report.Statistics 5 th Edition. Copenhagen, Denmark: ISWA Working Group on Thermal Treatment ofWaste; 2006. Available at:http://www.google.dk/search?q=International+Solid+Waste+Association+(ISWA).+Energy+from+Waste.+State+of+the+Art+Report.+Statistics+5th+Edition&sugexp=chrome,mod=15&sourceid=chrome&ie=UTF-8 (accessed 12.12).

Jungbluth, N., Chudacoff, M., Dauriat, A., Dinkel, F., Doka, G., Faist, E.M., Gnansounou, E.,Kljun, N., Schleiss, K., Spielmann, M., Stettler, C., Sutter, J. 2007. Life Cycle Inventories ofBioenergy. Uster (Switzerland): Swiss Centre for Life Cycle Inventories, ESU-services,Ecoinvent report No 17.

Juniper Consultancy Services Ltd., 2005. MBT: a Guide for Decision Makers: Processes-Policies-Markets. Available at: http://www.wmaa.asn.au/event-documents/2012skm/awt/or/Summary-Report.pdf (accessed 12.12).

Kirkeby, J.T., Hansen, T.L., Birgisdóttir, H., Bhander, G.S., Hauschild, M.Z., Christensen, T.H.,2006. Environmental assessment of solid waste systems and technologies: EASEWASTE.Waste Manage. Res. 24, 3-15.

Larsen, A.W., Vrgoc, M., Lieberknecht, P., Christensen, T.H., 2009a. Diesel consumption inwaste collection and transport and its environmental significance. Waste Manage. Res. 27, 652-

659.



43

Larsen, A.W., Merrild, H., Christensen, T.H., 2009b. Recycling of glass: accounting ofgreenhouse gases and global warming contributions. Waste Manage. Res. 27, 754-762.

Lund H., 2007. Renewable energy strategies for sustainable development. Energy 32, 912–919.

Manfredi, S., Christensen, T.H., 2009. Environmental assessment of solid waste landfillingtechnologies by means of LCA-modeling (EASEWASTE). Waste Manage. 29, 32–43.

Manfredi, S., Christensen, T.H., Scharff, H., 2009. Landfilling of waste: accounting ofgreenhouse gases and global warming contributions. Waste Manage. Res. 27, 825-836.

Manfredi, S., Christensen, T.H., Scharff, H., Joeri, J. 2010a. Environmental assessment of low-organic waste landfill scenarios by means of life-cycle assessment modelling (EASEWASTE). Waste Manage. Res. 28(2), 130-140.

Manfredi, S., Tonini, D., Christensen, T.H., 2010b. Contribution of individual waste fractions tothe environmental impacts from landfilling of municipal solid waste. Waste Manage. 30, 433-440.

Manfredi, S., Tonini, D., Christensen, T.H., 2011. Environmental assessment of differentmanagement options for individual waste fractions by means of life-cycle assessmentmodelling. Res. Cons. Recycl. 55, 995-1004.

Mathiesen, B.V., Lund, H., Karlsson, K., 2011a. 100% Renewable energy systems, climatemitigation and economic growth. Appl. Energy 88, 488–501.

Mathiesen, B.V., Lund, H., Hvelplund, F.K., Connolly, D., Bentsen, N.S., Tonini, D., Hansen,CEESA 100% Renewable Energy Scenarios towards 2050. Aalborg (Denmark): AalborgUniversity; 2011b. http://www.ceesa.plan.aau.dk (accessed 12.12).

Merrild, H., Damgaard, A., Christensen, T.H., 2009. Recycling of paper: accounting ofgreenhouse gases and global warming contributions. Waste Manage. Res. 27, 746-753.

Merrild, H., Larsen, A.W., Christensen, T.H., 2012. Assessing recycling versus incineration ofkey materials in municipal waste: The importance of efficient energy recovery and transportdistances. Waste Manage. 32, 1009-1018.

MMA, 2006. Medio Ambiente en España 2005. Ministerio de Medio Ambiente. Centro de publicaciones, Secretaria General Técnica. Avalaible at:http://www.marm.es/es/estadistica/temas/estadisticas-ambientales/index2010-10-29_23.19.09.1624.aspx (accessed 12.12).

MMA, 2010. Ministerio de Medio Ambiente. Perfil ambiental de España 2010. Centro de publicaciones. Secretaría Técnica. Available at:http://www.magrama.gob.es/es/ministerio/servicios/publicaciones/D_Gesti%C3%B3n_residuos

_tcm7-164926.pdf (accessed 12.12)

http://redirect.cvt.dk/?url=http%3A%2F%2Fdtu-ftc.cvt.dk%2Fcgi-bin%2Ffulltext%2Fsage%3Fpi%3D%2F10.1177%2F0734242X09104127%26key%3D140118788%26rfr_id%3Dinfo%3Asid%2Fdlib.dtu.dk%3ADTUDigitalLibrary


http://www.ceesa.plan.aau.dk/

http://www.ceesa.plan.aau.dk/





44

MITYC, 2011. Ministerio de Industria, Turismo y Comercio. La energía en España 2010.Secretaría General Técnica, Madrid. Available athttp://www.minetur.gob.es/energia/balances/balances/librosenergia/energia_espana_2012_2ed.pdf (accessed 12.12).

Montejo, C., Costa, C., Ramos, P, Márquez, M.C., 2010. Analysis of the presence of impropermaterials in the composting process performed in ten MBT plants. Bioresour. Technol. 101,8267-8272.

Montejo, C., Costa, C., Ramos, P, Márquez, M.C., 2011. Analysis and comparison of municipalsolid waste and reject fraction as fuels for incineration plants. Appl. Therm. Eng. 31, 2135-2140.

Meneses, M., Pasqualino, J., Castells, F., 2012. Environmental assessment of the milk life cycle:The effect of packaging selection and the variability of milk production data. J. Environ.

Manage. 107, 76-83.

Nielsen, M., Nielsen, O. and Thomsen, M., 2010. Emissions from decentralised CHP plants2007. Environmental project no. 07/1882. Project report 5 – Emission factors and emissioninventory for decentralised CHP production. National Environmental Research Institute, AarhusUniversity. Available at: http://www.dmu.dk/Pub/FR786.pdf (accessed 12.12).

Papageorgiou, A., Karagiannidis, A., Barton, J.R., Kalogirou, E., 2009. Municipal solid wastemanagement scenarios for Attica and their greenhouse gas emission impact. Waste Manag. Res.27, 928-937.

Riber, C., Petersen, C., Christensen, T.H., 2009. Chemical composition of material fractions inDanish household waste. Waste Manage. 29, 1251–1257.

Schmidt, A., Strömberg, K., 2006. "Genanvendelse i LCA - systemudvidelse." Rep. No.Miljønyt Nr. 81, Miljøstyrelsen. Available at: http://www2.mst.dk/udgiv/publikationer/2006/87-7052-165-4/pdf/87-7052-166-2.pdf (accessed 12.12).

Stegmann, R., 2010. Landfilling: MBP Waste Landfills, in Christensen T.H. (Ed) Solid WasteTechnology & Management. John Wiley & Sons, Ltd, pp. 788-799.

Stranddorf, H.K., Hoffmann, L., Schmidt, A., 2005. Påvirkningskategorier, normalisering ogvægtning i LCA – opdatering af udvalgte UMIP97-data. Danish Ministry of the Environment,Miljønyt n. 77. Available at: http://www2.mst.dk/udgiv/publikationer/2005/87-7614-572-7/pdf/87-7614-573-5.pdf (accessed 12.12)

Tetra Pak, 2004. Environmental Report 2004 Sweden. Available on line at:www.corporateregister.com/a10723/tps04-env-swe.pdf (accessed 12.12)

Tonini, D., Astrup, T., 2012a. Life-cycle assessment of a waste refinery process for enzymatictreatment of municipal solid waste. Waste Manage. 2012, 32, 165-76.

http://digitallibrary.dtu.dk/search?q=author:%22Meneses%2C%20Montse%22

http://digitallibrary.dtu.dk/search?q=author:%22Pasqualino%2C%20Jorgelina%22

http://redirect.cvt.dk/?url=http%3A%2F%2Fdtu-ftc.cvt.dk%2Fcgi-bin%2Ffulltext%2Felsevier%3Fpi%3D%2F0301%2F4797%2F0107000c%2F12002009.pdf%26key%3D366031111%26rfr_id%3Dinfo%3Asid%2Fdlib.dtu.dk%3ADTUDigitalLibrary


http://digitallibrary.dtu.dk/search?q=journaltitle=%22Journal%20of%20Environmental%20Management%22

http://www.corporateregister.com/a10723/tps04-env-swe.pdf

http://www.corporateregister.com/a10723/tps04-env-swe.pdf

http://digitallibrary.dtu.dk/search?q=journaltitle=%22Journal%20of%20Environmental%20Management%22



http://digitallibrary.dtu.dk/search?q=author:%22Pasqualino%2C%20Jorgelina%22

http://digitallibrary.dtu.dk/search?q=author:%22Meneses%2C%20Montse%22



45

Tonini, D., Astrup, T., 2012b. Life-cycle assessment of biomass-based energy systems: a casestudy for Denmark. Appl. Energy 99, 234-246.

Wenzel, H., Hauschild, M.Z, Alting, L., 1997. Environmental Assessment of Products.Methodology, tools and case studies in product development. Vol 1. Chapman & Hall, London,England.World Health Organization, WHO, 2006. Guidelines for Drinking-water Quality – Firstaddendum to third edition. Volume 1. ISBN 92 4 154696 4. WHO Press, 20 Avenue Appia,1211 Geneva 27, Switzerland.



VI

Potential for waste refineries in Europe

Tonini, D., Sanchez, V.M., Astrup, T.

Environmental Science and Technology, to be submitted



1

Potential for waste refineries in Europe

Davide Tonini *, Veronica Martinez Sanchez, Thomas Astrup

Department of Environmental Engineering, Technical University of Denmark, DTU –Building 115, 2800 Kgs. Lyngby, Denmark

* Corresponding author: [email protected] 0045 45251699

mailto:[email protected]






2

Abstract

Waste refineries are innovative

technologies allowing for separation of

degradable materials from mixed waste

into a bioliquid later digested for biogas

production. Recyclables can be sorted

from the solid fraction and the residual

combusted for additional energy

recovery: this promises optimized energy, material and resource recovery that can contribute

to reduce fossil fuel and resource depletion. This consequential life cycle assessment study

highlights the benefits and criticalities of waste refining against state-of-the-art waste

management: a total of 126 scenarios were evaluated. Overall, the greenhouse gas (GHG)

performance of waste refining was comparable with incineration and mechanical-biological

treatment: the relative GHG savings were within 15% difference. Waste refining performed

best with high organic content in waste, and may recover ca. 90% of the phosphorous in the

waste and increase the electricity production compared with incineration by 15-40%,

depending upon waste composition. However, nutrients recovery may come at the expenses

of additional toxics emission to soil. The sensitivity revealed that the choices about type of

energy substituted and waste composition are crucial: for example, bioreactor landfill

performed with the best GHG savings when heat substitution was not considered. Overall,

other environmental aspects than GHG seemed critical when comparing state-of-the-art waste

treatment technologies.



3

1. Introduction

In the endeavour of mitigating GHG emissions and reduce resource depletion increased

material, resource and energy recovery from mixed waste types (e.g. MSW) becomes

important. As demonstrated by recent studies (1-4), biofuels provision may represent a

challenge given the significant impacts associated with crops production, and thus energy

from residual substrates such as waste should be encouraged and optimized (5).

This acquires further relevance in the light of the European framework directive on

waste management (6) aiming at minimizing the amount of waste landfilled (especially with

respect to biodegradable materials) and maximizing instead the recovery of valuable

materials, resources and energy. Source-segregation of selected recyclables can contribute to

these goals; however, despite the efforts implemented in recent years in EU27, an average of

about 38% of MSW was disposed of in landfills, 20% incinerated and 42%

recycled/composted in 2009 (7). This first shows that source-segregation of recyclables is

still low in many EU geographical areas. Secondly, that a significant portion of the MSW (ca.

60%) ends up in the residual stream, hereafter named residual municipal solid waste (rMSW).

Additionally, the drastic limitations on landfilling of biodegradable organic materials

(with a limit of 35% on the ‘reference waste’ produced in 1995 by 2016) enforced by the (8)

yield high credits to organic waste separation. As for other materials, organic source-

segregation is possible and different techniques exist (9); however, the results are often far

from the expectations as segregation at the household and further pre-treatments prior to

biological conversion may lead to significant mass/energy/nutrients losses as highlighted in

recent studies (10).

In this perspective, technologies sorting materials, organic, and optimizing energy

recovery from mixed waste streams become attractive. For instance, mechanical-biological

treatment (MBT) plants typically use a combination of mechanical operations to separate the

organic fraction of the incoming mixed waste from the remaining materials; a share of these

are recovered while the residuals constitute refuse derived fuel (RDF) typically combusted

for energy recovery. The organic fraction is then biologically treated through composting or

anaerobic digestion (11). A disadvantage of such technology is that the stabilized organic

material (compost) is often contaminated by impurities preventing further recycling of

nutrients on land (12). In addition, scarce quality of the separated organic may reduce the

biogas yield during anaerobic digestion.

Emerging innovative technologies such as waste refineries promise improved organic,materials and energy recovery (13). The waste refinery generates two main products from the



4

mixed waste: i) a bioliquid derived from enzymatic liquefaction of degradable materials

(organic, paper and cardboard) and ii) a residual solid (non-degradable materials). Metals and

plastic can be further separated from this stream. The bioliquid can be digested to produce

biogas or ethanol, or co-fired in power plants. The residual solid can be combusted allowing

for additional energy recovery. Further, bioliquid/biogas are storable energy carriers and can

be used in a variety of applications (e.g. gas engines, gas turbines, boilers, upgrading for use

in transport sector, conversion to ethanol, etc.) providing additional flexibility to the whole

energy system. This is important in the perspective of future energy systems based on high

penetration of fluctuating energy sources such as windenergy, hydropower, photovoltaic, etc.

(3, 14-18).

In the perspective of optimizing energy, materials and resources recovery from waste

the objectives of this study are: I) to assess the environmental and energy performance of

waste refining scenarios against a number of references representing state-of-the-art MSW

management; these involved incineration, landfilling and MBT. II) To assess the

environmental and energy relevance of organic source-segregation and its overall potentials.

III) To evaluate the importance of the waste composition on the LCA results. IV) To identify

the framework conditions (e.g. waste composition, efficiency of energy recovery

technologies, etc.) whereby waste refineries can be preferred over the remaining

technologies.



5

2. Methodology

2.1 Goal, scope and functional unit

The environmental assessment presented in this study was performed using consequential life

cycle assessment (19, 20). The functional unit was "treatment of 1 tonne (1 t = 1 Mg) of

municipal solid (wet) waste, i.e. MSW". The geographical scope of the study was Europe.

Accordingly, the waste composition was based on recent studies investigating European

waste (21, 22). The chemical composition was based on (21). The waste treatment

technologies were modelled as state-of-the-art technologies which are or are going to be

established in the future years.

All environmental impacts (resources consumption, emissions to air, soil and water)

were included for a time horizon of 100 years. The ‘zero burden’ approach was applied: all

upstream emissions associated with generating the waste were omitted from the LCA (i.e. the

waste as such was assumed to carry no-impacts). Downstream utilization of recovered energy

and recyclables were credited the LCA scenarios by system expansion into the energy and

industrial sectors (avoided production of energy and virgin materials). The boundary of the

system was set at the household and included collection, transportation, treatment, disposal of

residues (e.g. incineration ashes, stabilized organic material, etc.), recycling of materials and

application on land of compost from anaerobic digestion. The environmental impacts

associated with the construction and demolitions of facilities were not included.

2.2 Impact assessment

The assessment was carried out according to the LCA method EDIP 1997 (23). The

following impact categories were included in the assessment: global warming (GW),

acidification (AC), nutrient enrichment (NE), ecotoxicity in water chronic (ETwc), human

toxicity via water (HTw) and human toxicity via soil (HTs). Additionally, an impact category

named “phosphorous resource saving” (P res) was included in order to reflect the benefits

associated with use on land of compost produced from anaerobic digestion and post-

composting.

2.3 LCA scenarios modeling and system boundary

The systems assessed considered two different sets of waste composition (namely a and b) to

include the variety intrinsically connected to MSW and a number of waste managementscenarios differing for the treatment of the residual waste (that is, left-over after source-



6

segregation of selected recyclables) and of the organic waste (i.e. vegetable and animal waste,

kitchen tissues and wood). A total of 42 scenarios have been assessed (the different

combinations are illustrated in Figure 1). For all technologies, the energy produced was

assumed to be used for combined heat and power (CHP) production (with the exception of

one scenario where transport fuel was substituted) substituting marginal electricity and heat

production (or transport fuel). Of the fossil fuels, coal and natural gas represent the two ends

of the range with respect to CO 2 emissions per combustion unit of fossil fuel energy. At a

European level, these are also the fossil fuels likely to react to increased or decreased energy

production from waste (see supporting information, SI). For the baseline calculations, coal

was assumed as marginal for electricity production. This assumption was tested in the

sensitivity analysis by substituting electricity from natural gas. For heat, three scenarios

representative of Western/Southern, Eastern and Northern Europe heat markets were

assessed: this involved substitution of heat produced from: 1) natural gas boilers (baseline),

2) coal boilers and 3) coal-fired CHP plants (district heating). For transport fuels, gasoline

was assumed as marginal in the baseline. This assumption was tested in the sensitivity

analysis by substituting diesel fuel.

The compost produced from biological treatment of source-segregated organic waste

was used as a fertilizer which avoided marginal mineral N, P and K fertilizers to be produced

and used, based on the content of N, P and K of the compost. The marginal N, P and K

fertilizers considered were calcium ammonium nitrate, diammonium phosphate and

potassium chloride, respectively, conformingly with (24, 25) (see SI).

2.3.1 Waste composition

Two different sets of waste composition were used for the assessment (summarized in Table

S1): ( a ) a Danish waste composition (21) and ( b) a Spanish waste composition (22). The

principal difference between the two datasets was the share (% of the total) of organic waste:

in (22) this added up to about 59%, whereas in (21) to approximately 34%. The share of

paper and cardboard materials amounted to 30% in (21) and to 14% in (22). The two waste

compositions represent a variety often encountered (26-28).

2.3.2 Focus on the scenarios

Source-segregation of recyclables and organic waste

All the assessed waste management scenarios were modelled by assuming household source-segregation of selected recyclable materials namely: I) ferrous metals, II) aluminium, III)



7

plastic (hard plastic, i.e. HDPE and HDPP, plastic bottles, i.e. PET and soft plastic), IV)

paper and cardboard (dirty paper and cardboard excluded) and V) glass. As separation

efficiencies, typical source-segregation efficiencies were assumed: 75% for ferrous metals

and aluminium, 50% for plastic, 80% for paper and cardboard and 75% for glass. These

values represent average segregation efficiencies in different types of collection systems

ranging from full-service (door to door) to joint collection points (e.g. in apartments) as

illustrated in (29). The segregated materials were sent to facilities for further recycling.

With respect to the organic waste (i.e. vegetable and animal waste, kitchen tissues and

wood) three alternatives were investigated: a scenario with no source-segregation ( 0), i.e. all

organic waste ended up in the residual waste, a “realistic” scenario with 70% source-

segregation efficiency ( I ) and a scenario with 100% efficiency ( II ); the latter one was

assessed to evaluate the potential maximal benefits associated with 100% organic waste

segregation efficiency. Note that the considered 70% efficiency for scenario ( I ) falls in the

middle of the range provided in (29). Only vegetable, animal waste, kitchen tissues and wood

were considered for organic source-segregation to avoid contamination with impurities.

As source-segregation was not 100% efficient, recyclable materials and organic were

still present in the rMSW. An overview of the composition of the rMSW is presented in

Table S1.

Treatment of the residual waste

A number of waste management scenarios were considered for the treatment of the residual

waste: (1) incineration with CHP production ( INC ), (2) conventional landfilling with biogas

flaring ( CLF ), (3) landfilling in bioreactor with electricity production from the collected

biogas ( BLF ), (4) mechanical-biological treatment with anaerobic digestion for CHP

production and post-composting of the digested material ( MBT AC ), (5) mechanical-

biological treatment with direct composting of the organic material ( MBT DC ), (6) waste

refining with anaerobic digestion of the bioliquid and further biogas conversion in gas

engines for CHP production ( WR GE ) and (7) waste refining with anaerobic digestion of the

bioliquid and further biogas upgrading to methane for use as transport fuel ( WR TF ).

Overall, 126 waste management scenarios were assessed (Figure 1). These resulted

from the combination of 2 (waste compositions, a and b), 3 (management scenarios for

organic waste, i.e. 0, I and II ), 7 (management scenarios for residual waste) and 3 (energy

system scenarios), i.e. 2 x 3 x 7 x 3 = 126.



8

An overview of the LCA system boundary is presented in Figure S1-S14 for the case

of scenarios ( 0) and ( I ) given waste composition a (baseline energy system). Detailed

information on waste materials and energy balances for the case II (100% organic source-

segregation) and for the scenarios with waste composition b can be found in the SI.

Figure 1. Overview of the scenarios assessed in the LCA.

2.4 LCI of waste treatment technologies

2.4.1 The waste refinery

The waste refinery process was based on a pilot-scale facility established in Copenhagen,

Denmark (SI). The waste refinery aims at producing two main products from the incoming

mixed MSW: i) a bioliquid (i.e. slurry composed of enzymatically liquefied organic, paper

and cardboard) and a residual solid (i.e. non-degradable waste materials).The refinery process consisted of two reactors: in the first reactor the waste was

heated by injection of hot water to about 75 °C for approximately 0.5-1 hours, then cooled to

about 50-55 °C before entering the second reactor. In the second reactor enzymes were added

(about 5 kg t -1 MSW) resulting in hydrolysis and break-down of bonds in the organic

materials thereby essentially suspending organic materials in a liquid phase (30). The

retention time was about 10-16 hour. After the second reactor, the liquid phase was separated

from the remaining solids by a vibrating sieve. Further, another vibrating sieve separated the

liquid phase into a bioliquid and a solid “fluff” (containing materials such as cotton and



9

textiles, but also glass and plastic pieces). The bioliquid consisted primarily of suspended

organic matter (food waste, paper and cardboard), while the solid fraction mainly consisted of

non-degradable materials such as plastic, metals, textiles, soil, ceramics, etc .

The bioliquid can be exploited for biogas production (option considered in this study),

co-combusted in coal-fired power plant or utilized for producing ethanol. The solid fraction

can undergo further sorting for metal and plastic recovery. The remaining residual solid can

be combusted for energy recovery. Overall, electricity and heat consumption were estimated

to 25 kWh t -1 MSW and 490 MJ t-1 MSW. Additional information on the process and

bioliquid properties can be found in the SI.

2.4.2 Other technologies and processes

A detailed description of the remaining technologies (i.e., incineration, conventional

landfilling, landfilling in bioreactor, mechanical-biological treatment plant, biological

treatment, use on land, collection and transport, etc.) used in the LCA can be found in the SI.

2.5 Sensitivity and uncertainty analysis

Sensitivity and uncertainty analysis was addressed on two levels: i) scenario uncertainties and

ii) parameters uncertainties. Model uncertainties (which refer to uncertainties in the LCA

methodology, in the equations used for modelling, etc.) were not addressed. The reason for

this was that the uncertainty of the LCA methodology equally applied to all the selected

scenarios. Further, the modelling equations were basic mass/energy balances which do not

require an uncertainty assessment.

With respect to (i) a number of sensitivity analyses were performed on the assessed

scenarios. These included: (1) natural gas as marginal for electricity production (instead of

coal as in the baseline); (2) no heat recovery (instead of heat recovery as in the baseline); (3)

diesel as marginal for transport (instead of gasoline); (4) landfilling of the compost from

bioliquid digestion (instead of use on land as in the baseline). For (1), (2) and (3) the focus

was on GW. For (3), the focus was on the environmental categories: NE, HTw and HTs;

landfilling of the compost from bioliquid was here assessed as an example of alternative

disposal method to use on land in order to avoid leaching and soil contamination. Other

alternatives may be of interest (see section 3) such as co-digestion or incineration. Each of

these changes was individually tested to assess the influence of the individual change on the

overall LCA results.



10

Furthermore, the influence of relevant parameters uncertainty (ii) on the ranking provided by

the LCA results was tested for two selected scenarios ( INC 0 and WR GE 0 ) by individually

varying selected parameters. This allowed identifying the conditions (e.g. an interval of

values or a threshold) where one management scenario allowed for more environmental

savings than the other. The focus was on the category GW.


The results of the LCA are presented in Figure 2-3 respectively for non-toxic (along with P

resource) and toxic impact categories (coal as marginal for electricity and natural gas as

marginal for heat). Figures S19-S20 show the results for the case of natural gas as marginal

for electricity. The results are expressed as characterized impact potentials per tonne of wet

waste. Impacts/savings for the individual scenarios were obtained by subtracting the avoided

impacts (negative values in the figures) from the induced impacts (positive values). Any net

value below the zero axes indicates an environmental improvement compared with the fossil

fuel reference (in which electricity is provided by coal and materials are produced from virgin

resources). The ‘triangle’ indicates the savings/impacts associated with treatment of rMSW

only. The ‘circle’ indicates the savings/impacts associated with treatment of rMSW plus

source-segregated organic. The ‘square’ indicates the total savings/impacts relative to 1 t

MSW (i.e., including savings associated with recycling of source-segregated aluminium,

ferrous metals, paper, plastic and glass). The focus of the discussion is on the management of

rMSW and organic waste as the results associated with source-segregation of aluminium,

ferrous metals, plastic, paper and glass are the same for all a (or all b) scenarios. Additional

details on energy and waste materials balance is provided in the SI. The results of the

sensitivity analyses are also detailed in the SI.

3.1 Environmental performances without organic source-segregation (scenarios 0)

Waste refining performed comparably to incineration and MBT (involving anaerobic

digestion and CHP) on GW and AC, provided use of biogas for CHP. Among these scenarios,

the absolute relative difference in GHG savings was small (about 2-15%, see Table S9). The

final GW ranking was dictated by the waste composition ( a and b): with low-organic waste

composition incineration performed best; viceversa, with high-organic waste composition

waste refining with CHP achieved the largest GHG savings. Scenarios without energy

recovery ( CLF and MBT DC ) were always the worst on GW and AC. These results reflectedthe outcome of the energy balance (Figure S16) where incineration and waste refining with



11

CHP performed best: the latter showed the highest electricity recovery for both a and b while

incineration achieved by far the best heat recovery.

The waste refining scenarios involving upgrading of the biogas to transport fuel

showed lower performances GHG-wise. This was primarily a consequence of the lower CO 2

emission factor associated with the fuel substituted (gasoline) compared to that of coal. In

fact, when natural gas was the marginal (Figure S20) using biogas for CHP or for transport

was comparable (see 3.4). Additionally, significant savings were found on ETwc as a result

of avoided PAH emissions from gasoline production (Figure 3). This acquires further

importance when considering that provision of transport biofuels in future sustainable energy

systems might be a challenge and induce energy crops production (4). Considered the

possible drawbacks associated with these (e.g., iLUC), the use of waste for transport biofuel

production may represent a valuable alternative.

The recovery of P was maximized in the waste refining scenarios ( WR GE and WR

TF ) thanks to the high potential for organic recovery in the form of bioliquid. This, however,

came at the expenses of additional eutrophication (NE) and toxic impacts (HTs and HTw):

the first because of enzymes use (with corresponding eutrophication impacts, see also (13))

and the second because of increased heavy metals transfer to the bioliquid (and relative

digestate). Similar findings were in (13) where, however, the digestate did not undergo post-

composting inducing even higher N-leaching, conformingly with the findings of other studies

(e.g., (31)).



12

Figure 2. LCA results for the non-toxic environmental categories (coal as marginal for electricity andnatural gas as marginal for heat). The ‘triangle’ indicates the savings/impacts associated with thetreatment of rMSW only. The ‘circle’ indicates the savings/impacts associated with treatment ofrMSW plus source-segregated organic. The ‘square’ indicates the overall savings/impacts relative to 1t MSW.

3.2 Effects of organic source-segregation (scenarios I and II )

Overall, as shown by Figure 2-3, organic source-segregation would be meaningless in waste

refining scenarios as here the all idea is to recover the organic in form of bioliquid, so that

source-segregation would not be needed. Instead, from a GHG perspective source-

segregation did benefit ( INC ) and MBT scenarios ( MBT AC ) in the case of high-organicwaste composition ( b): both performed comparably to waste refining with CHP ( WR GE ),



13

provided source-segregation of the organic waste (scenarios I , see also Table S9). This was a

result of increased energy recovery from the rMSW sent to incineration as an effect of its

higher LHV after organic segregation. Also, those scenarios not performing energy recovery

from the rMSW ( CLF and MBT DC ) significantly increased their total GW and AC savings

thanks to the energy recovery associated with biological conversion of segregated organic

waste. This was an expected finding as confirmed by previous studies ((26) among the

others). These benefits did not apply to NE where the emissions of NH 3 and NO 3- occurring

during the biological process and subsequent use on land determined an overall lower (or

comparable) saving compared with the scenarios 0.

Based on this, it is straight-forward that higher organic source-segregation efficiency

(with the potential maximum shown in Figure 2-3 for scenarios II ) would primarily benefit

those scenarios not performing energy recovery from rMSW ( CLF and MBT DC ).

Incineration ( INC ) and MBT with combined treatment ( MBT AC ) would benefit in case of

high-organic waste composition ( b), as earlier mentioned.

With the exception of waste refining scenarios, in the remaining P savings were ca.

56% of the potential maximum ( II ) when realistic source-segregation was considered ( I ), due

to the losses occurring during source-segregation and mechanical pre-treatment.

3.3 The influence of the waste composition

The waste composition affected the ranking of the scenarios and the magnitude of the

savings. This primarily regarded GW, the toxicity categories (HTw and HTs) and P resource.

As mentioned in 3.1, the waste composition was crucial in determining the ranking of the

scenarios on GW when organic source-segregation was not considered. The outputs of waste

refining, in fact, are a function of organic, paper and cardboard content and therefore may be

significantly affected by changes in waste composition, including water content of the

materials. Similarly, MBT and incineration performances are affected by changes in

composition and water content. This is highlighted in Figure S16 where an energy balance is

presented: for example, the electricity production in waste refining with CHP ( WR GE ) was

considerably higher than incineration ( INC ) in b (increase of 40%) compared with a (increase

of 15%) provided that organic source-segregation was not implemented ( 0). The reason for

the higher recovery in b was the increased amount of rMSW as well as organic waste (as

share of the total rMSW) in b as evident from Table S1. Conversely, the heat production in

the waste refining CHP scenarios was decreased compared with direct incineration (-19% inb and -30% in a). This, along with the environmental impacts caused by the pre-treatment



14

itself (energy and enzymes consumption) was the reason why incineration ( INC ) was

environmentally better than waste refining with CHP ( WR GE ) on GW and AC in the case a.

This, however, was not true in the case b where the higher share of organic waste induced

correspondingly higher biogas production which largely compensated for the pre-treatment

impacts.

On HTw the differences in waste composition caused opposite results for waste

refining scenarios in a and b case: in this category, in fact, the final result is a balance

between Hg and metals emissions from use on land and saved Hg and metals emissions from

avoided fertilizers production. A net impact was quantified in the case a (where the induced

impacts associated with use on land exceeded the savings associated with the avoided Hg and

metals emissions connected with mineral fertilizer production) whereas a net saving resulted

in b as a consequence of the larger amount of mineral fertilizer substituted (more compost

was produced) as well as of the lower concentration of Hg and metals in the bioliquid (due to

a lower share of bioliquid derived from materials other than organic waste, see Table S1). It

can be noted that in the case a the HTw impact did not decrease as an effect of organic

source-segregation ( I and II ); this was an expected result as Hg and heavy metals are not

transferred to bioliquid from organic waste but rather from paper, cardboard, cat litter,

cartons, metals, etc. These waste fractions, even with organic source-segregation, would still

be treated in the waste refinery (as part of the rMSW) inducing similar substance transfer.

As it can be noted from Figure 2-3, the ranking of the scenarios for Pres and HTs

were the same in a and b. However, the magnitude of the savings increased in the case b as a

consequence of the larger share of organic in the waste composition that induced higher

benefits from use on land of the compost, similarly to the previous results for HTw.

The results of the comparison between direct incineration ( INC 0 ) and waste refining

with CHP ( WR GE 0 ) highlighted the crucial importance of the waste composition in

determining under which conditions one alternative is favourable over the other, from a GHG

perspective. From Figure S21 it is evident that with low-organic waste composition ( a) only

low electricity efficiency of incineration (<15%) and/or high electricity efficiency of biogas

conversion (>50%) may favour waste refining over incineration. Viceversa, with high-

organic waste composition ( b), the electricity efficiency of incineration should be greater than

32% to overcome the waste refinery GHG performance.



17

electricity conversion efficiency should be greater than 50% for waste refining to achieve the

same GHG savings of state-of-the-art incineration (alternatively, the latter should have very

low efficiency, i.e. < 15%).

Overall, state-of-the-art incineration, MBT, landfilling (bioreactors) and waste

refining may provide comparable GHG savings. In the light of this, if a decision is to be

made, relative global warming performances appear less important than other environmental

categories and issues that may be instead critical discriminating one alternative in favour of

another. This is the case of use on land-related impacts and P resource savings.

Supporting Information (SI)

Additional information on: marginal energy technologies and fertilizers, LCA process flow

diagrams, waste composition, LCI of waste treatment technologies, waste materials balance,

energy balance, LCA results for the different scenarios assessed as well as sensitivity and

uncertainty analyses is available free of charge via the Internet at http://pubs.acs.org.

Acknowledgements

Financial support to this study was provided by the Research Grant EUDP 304701 from

Danish Energy Agency (DEA) as well as Technical University of Denmark (DTU).



18

References

1. Searchinger, T.D.; Heimlich, R.; Houghton, R.A.; Dong, F.; Elobeid, A.; Fabiosa, J.; Tokgoz, S.;Hayes, D.; Yu, T.H. Use of U.S. croplands for biofuels increases greenhouse gases through emissionsfrom land-use change. Science 2008 , 319 , 1238-1240.

2. Searchinger, T.D. Biofuels and the need for additional carbon. Environmental Research Letters 2010 , 5, 024007-024007.

3. Tonini, D. and Astrup, T. Life-cycle assessment of biomass-based energy systems: A case study forDenmark. Appl. Energy 2012 , 99 , 234-246.

4. Tonini, D.; Hamelin, L.; Wenzel, H.; Astrup, T. Bioenergy Production from Solid Biomass: aConsequential LCA of 12 Bioenergy Scenarios including Land Use Changes. Environ. Sci. Technol .2012 , 46(24) , 13521-13530.

5. Steubing, B.; Zah, R.; Ludwig, C.; Heat, electricity, or transportation? The optimal use of residualand waste biomass in Europe from an environmental perspective. Environ. Sci. Technol . 2012 , 46 ,164-171.

6. Directive 2008/98/EC of the European Parliament and of The Council of 19 November 2008 onwaste and repealing certain Directives. Official journal of the European Union 2008, L312/3;http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2008:312:0003:0030:en:PDF (accessedJanuary 30, 2013).

7. Europe in figures - Eurostat yearbook 2011. Eurostat: Luxembourg, European Union, 2011;http://epp.eurostat.ec.europa.eu/portal/page/portal/publications/eurostat_yearbook_2011 (accessedJanuary 30, 2013).

8. Directive 1999/31/EC of the European Council of 26 April 1999 on the landfill of waste. OfficialJournal of the European Union 1999, L 182/1;http://www.central2013.eu/fileadmin/user_upload/Downloads/Document_Centre/OP_Resources/Landfill_Directive_1999_31_EC.pdf (accessed January 30, 2013).

9. Bernstad, A. and Jansen, J. la Cour. Separate collection of household food waste for anaerobicdegradation – Comparison of different techniques from a systems perspective. Waste Manage . 2012 ,

32, 806-815.

10. Bernstad, A.; Malmqvist, L.; Truedsson, C.; Jansen, J. la Cour. Need for improvements in physical pre-treatment of source-separated household food waste. Waste Manage . 2012 . DOI:10.1016/j.wasman.2012.06.012.

11. Bilitewski, B.; Oros, C.; Christensen, T.H. Mechanical Biological Treatment, In Solid WasteTechnology & Management ; Christensen T.H. Ed.; John Wiley & Sons, Ltd: 2010; pp. 628-638.

12. Montejo, C.; Ramos, P.; Costa, C.; Marquez, M.C. Analysis of the presence of improper materialsin the composting process performed in ten MBT plants. Bioresour. Technol . 2010 , 101 , 8267-8272.

13. Tonini, D. and Astrup, T. Life-cycle assessment of a waste refinery process for enzymatictreatment of municipal solid waste. Waste Manage . 2012 , 32, 165-176.



19

14. Mathiesen, B.V.; Lund, H.; Karlsson, K. 100% Renewable energy systems, climate mitigation andeconomic growth. Appl. Energy 2011 , 88, 488-501.

15. Lund, H. and Mathiesen, B.V. Energy system analysis of 100% renewable energy systems: Thecase of Denmark in years 2030 and 2050. Energy 2009 , 34, 524-531.

16. Lund, H. Renewable energy strategies for sustainable development. Energy 2007 , 32, 912-919.

17. CEESA 100% Renewable Energy Scenarios towards 2050 . Aalborg University. Aalborg,Denmark, 2011;http://www.ceesa.plan.aau.dk/digitalAssets/32/32603_ceesa_final_report_samlet_02112011.pdf(accessed January 30, 2013).

18. Mathiesen, B.V.; Lund, H.; Noergaard, P. Integrated transport and renewable energy systems.Utilities Policy 2008 , 16 , 107-116.

19. Environmental Management-Life Cycle Assessment-Principles and Framework . 2006-07-01, 2 nd ed. 2006; ISO 14040.

20. Environmental Management-Life Cycle Assessment-Requirements and Guidelines . 2006-07-01, 1 st ed. 2006; ISO 14040.

21. Riber, C.; Petersen, C.; Christensen, T.H. Chemical composition of material fractions in Danishhousehold waste. Waste Manage . 2009 , 29, 1251-1257.

22. Montejo, C.; Costa, C.; Ramos, P.; Marquez, M.C. Analysis and comparison of municipal solidwaste and reject fraction as fuels for incineration plants. Appl. Therm. Eng . 2011 , 31, 2135-2140.

23. Wenzel, H.; Hauschild, M.Z.; Alting, L. Environmental Assessment of Products - Volume 1: Methodology, Tools and Case Studies in Product Development ; Kluwer Academic Publishers: Norwell, MA., 1997.

24. Hamelin, L.; Wesnaes, M.; Wenzel, H.; Petersen, B.M. Environmental Consequences of FutureBiogas Technologies Based on Separated Slurry. Environ. Sci. Technol . 2011 , 45, 5869-5877.

25. Hamelin, L.; Joergensen, U.; Petersen, B.M.; Olesen, J.E.; Wenzel, H. Modelling the carbon andnitrogen balances of direct land use changes from energy crops in Denmark: a consequential life cycleinventory. GCB Bioenergy 2012 , 4(6) , 889-907.

26. Christensen, T.H.; Simion, F.; Tonini, D.; Moeller, J. Global warming factors modelled for 40generic municipal waste management scenarios. Waste Manage. Res . 2009 , 27 , 871-884.

27. Troschinetz, A.M. and Mihelcic, J.R. Sustainable recycling of municipal solid waste in developingcountries. Waste Manage . 2009 , 29, 915-923.

28. Asase, M.; Yanful, E.K.; Mensah, M.; Stanford, J.; Amponsah, S. Comparison of municipal solidwaste management systems in Canada and Ghana: A case study of the cities of London, Ontario, andKumasi, Ghana. Waste Manage . 2009 , 29, 2779-2786.

29. Christensen, T.H. and Matsufuji, Y. Source Segregation and Collection of Source-Segregated

Waste, In Solid Waste Technology & Management ; Christensen T.H. Ed.; John Wiley & Sons, Ltd:2010; pp. 296-310.



20

30. Jensen, J.W.; Felby, C.; Jørgensen, H.; Roensch, G.Oe.; Noerholm, N.D. Enzymatic processing ofmunicipal solid waste. Waste Manage . 2010 , 30, 2497-2503.

31. Boldrin, A.; Neidel, T.L.; Damgaard, A.; Bhander, G.S.; Moeller, J.; Christensen, T.H. Modellingof environmental impacts from biological treatment of organic municipal waste in EASEWASTE.

Waste Manage . 2011 , 31, 619-630.

32. Bekendtgørelse nr. 1650 af 13 December 2006 om anvendelse af affald til jordbrugsformaal(Slambekendtgoerelsen). Public law, 2006;https://www.retsinformation.dk/Forms/R0710.aspx?id=13056 (accessed January 30, 2013).

33. Vergara, S.E.; Damgaard, A.; Horvath, A. Boundaries matter: Greenhouse gas emissionreductions from alternative waste treatment strategies for California's municipal solid waste. Resour.Conserv. Recycl . 2011 , 57 , 87-97.

34. Manfredi, S. and Christensen, T.H. Environmental assessment of solid waste landfilling

technologies by means of LCA-modeling. Waste Manage . 2009 , 29, 32-43.35. Manfredi, S.; Tonini, D.; Christensen, T.H.; Scharff, H. Landfilling of waste: accounting ofgreenhouse gases and global warming contributions. Waste Manage. Res . 2009 , 27 , 825-836.



The Department of Environmental Engineering (DTU Environment) conductsscience-based engineering research within four sections:Water Resources Engineering, Urban Water Engineering,

Residual Resource Engineering and Environmental Chemistry & Microbiology.

The department dates back to 1865, when Ludvig August Cold ing, thefounder of the , gave the first lecture on sanitary engineering asresponse to the cholera epidemics in Copenhagen in the late 1800s.

department

Davide Tonini PhD Thesis FULL Version for ORBIT.pdf

Documents

Transcript of Davide Tonini PhD Thesis FULL Version for ORBIT.pdf