Life-cycle analysis a look into the key parameters affecting life-cycle CO 2 emissions of passenger cars

Transcription

1 Life-cyce anaysis a ook into the key parameters affecting ife-cyce CO 2 emissions of passenger cars Introduction The genera framework and guideines for a ife-cyce assessment (LCA) are defined in ISO 14044:2006. This standard defines the genera principes of a methodoogy used to assess the environmenta impact of different products, from the extraction of the raw materias, through their use and finay recycing or disposa of the end-of-ife and waste materias. Figure 1 shows how an LCA woud appy to vehices. When appied to the CO 2 emissions associated with a vehice, the LCA incudes the we-to-whees (WTW) emissions generated in the production and consumption of different fues, as we as the CO 2 emissions associated with the production and disposa of the vehice, incuding the batteries in the case of eectric powertrains. This artice presents the basis of a Concawe anaysis comparing the ife-cyce CO 2 emissions of battery-eectric vehices (BEVs) and interna combustion engine (ICE) vehices, and investigates the impact of a range of key parameters on tota CO 2 emissions. Figure 1: LCA appied to vehices a bigger picture The aforementioned ISO 14044:2006 standard, despite providing genera guideines, is not specific enough to ensure a singe and homogeneous methodoogy for conducting an LCA for fues and powertrains. As an exampe, different methodoogies to assess the energy consumption associated with battery manufacture can be found in recent iterature, eading to very different resuts. Beyond the different approaches used, there is aso a need to access more detaied and pubic data from manufacturers to Concawe Review Voume 27 Number 1 Juy

2 ensure the robustness and representativeness of the fina resuts. Whie these imitations prevent the LCA methodoogy from being used more widey, it is commony agreed that such an anaysis provides the key eements to perform a fu technica comparison of the environmenta impact of distinct energy/propusion aternatives such as interna combustion engine (ICE) vehices and battery eectric vehices (BEVs). Concawe has used the LCA methodoogy to assess the CO 2 emissions associated with different fue and powertrain combinations. This artice presents the basis of that anaysis, and investigates the impact that certain key parameters may have across the whoe ife cyce of a vehice. The resuts and figures incuded are initia estimates based on reevant externa pubications and on Concawe s own interna research. Concawe is wiing to engage with externa stakehoders to assist in defining a standard LCA methodoogy to be appied to fues and powertrains in the future. Using LCA to assess CO 2 emissions from passenger cars Basis When comparing different eectric vehices with conventiona powertrains, and the reative impacts associated with the energy or fue generation, materias extraction, and manufacturing and production phases, the contributions to the tota ife-cyce emissions are distinct. The key parameters affecting the resuts are summarised in Tabe 1. Tabe 1: Key parameters affecting vehice LCA INTERNAL COMBUSTION ENGINE (ICE) BATTERY ELECTRIC VEHICLE (BEV) Non-road factors * * The % of non-use emissions of different powertrains may vary from 20 30% for ICEs, and from 30 70% for BEVs depending on the key parameters identified. [1] Vehice cass (e.g. A, B, C) Drivetrain materias (stee, auminium) Production of the fue Vehice cass (e.g. A, B, C) Drivetrain materias (copper) Battery production Type (materias) Size / Range Ce production country Battery assemby area CO 2 estimation mode Energy use for eectrica generation (we-to-tank) Carbon intensity of the eectricity mix Road factors Vehice use Lifetime (years) Tota kiometres driven Use (urban, rura, etc.) Type of fue (e.g. petro/diese) Fue consumption Quantity (/100 km) Drive cyce (NEDC, WLTP, RDE) Vehice use Lifetime (years) Tota kiometres driven Use (urban, rura, etc.) Unit consumption Quantity (kwh/100 km) Drive cyce 18 Concawe Review Voume 27 Number 1 Juy 2018

3 Tabe 2: Concawe LCA simpe modeing too main inputs PARAMETER VALUE COMMENT SOURCE Driving distance 150,000 km To ensure that no battery repacement is required. Concawe estimate. Embedded emissions (battery manufacturing) 150 kg CO 2 /kwh Lithium ion battery (NMC). Top-to-bottom approach. Defaut vaue assumed constant regardess of the battery size (simpification). Average vaue from IVL report. Energy use for battery manufacturing: MJ/ KWh. Embedded emissions, vehice manufacturing 4 t CO 2 /vehice (Cass B) 5 t CO 2 /vehice (Cass C) 7 t CO 2 /vehice (Cass D) Generic vaues (ack of data per individua vehice). Based on NTNU. EU eectricity mix (ow votage, incuding osses) 350 g CO 2 /kwh ow votage (2016). Preiminary estimate (see comments and sources) EU LV eectricity mix (EU-28) preiminary estimate based on adjusted IEA data (WEO JRC/JEC methodoogy incuding upstream emissions and osses). Reference: 2013: 447 g CO 2 /kwh (JRC detaied anaysis). 2016: 300 g CO 2 /kwh (IEA WEO) EU eectricity generation mix (HV without upstream emissions). 2016: Concawe preiminary estimate (subject to change once the updated JRC work is pubic).* 2013 vaue: JRC detaied anaysis. 2 IEA WEO Eectricity/fue consumption Variabe Specific for each mode considered. OEM brochures. Rea-driving emissions (RDE) adjustment factor 1.4 Correction factor used to upift fue consumption from NEDC (New European Driving Cyce) to RDE vaues. Concawe estimate. Conservative vaue aigned with other sources (EU Commission modeing). Charging osses 10% Defaut vaue aigned with 90% on-board battery charger efficiency (Conservative vaue. Fast charging not incuded). Based on NTNU, Ricardo data. End-of-ife (EOL) emissions 0.5 t CO 2 /vehice (BEVs) 0.4 t CO 2 /vehice (ICEs) Battery 20% of tota EOL emissions for BEVs. Based on NTNU data. * JRC is currenty working on a paper cacuating the 2015 LV CO 2 intensity vaue using the most recent IEA statistic data issued with 2 years of deay (detaied methodoogy). [2] Concawe Review Voume 27 Number 1 Juy

4 The infuence of some of these parameters have been expored by Concawe, and the initia resuts are presented beow based on information currenty avaiabe in externa sources such as NTNU, [3] IVL, [4] European Commission data adapted from BMVI [5] as we as different OEMs brochures for individua vehices. Based on these sources, Concawe has deveoped an LCA mode to expore different countryspecific scenarios and run a sensitivity anaysis on the key parameters across a vehice segments. This mode is intended to be a ive too to conduct periodic assessments as new data become pubicy avaiabe. The main inputs used are summarised in Tabe 2 on page 19. Sensitivity anaysis Manufacturing stage: battery type and size, and eectricity mix used Different pubished LCA studies show a wide variabiity in the embedded CO 2 emissions of BEVs reated to the battery manufacturing process. Generay, these show that BEVs have higher embedded greenhouse gas (GHG) emissions than equivaent gasoine and diese ICEs primariy due to: Methodoogica factors such as the chosen ife-cyce inventory (LCI) database or the LCA methodoogy used (top-down or bottom-up approach). The seection of the manufacturing cacuation method is one of the main causes of the discrepancies in embedded emissions found in iterature. Whie top-down studies aocate energy use based on information about the individua process, the bottom-up approach aggregates data from each individua activity incuding energy consumption from utiities and additiona auxiiary processes. The top-down approach causes higher greenhouse gas emissions and cumuative energy demand. Battery pack size and chemistry/technoogy used; this aso determines the energy density of the package. CO 2 intensity of the energy mix used during the manufacturing and assemby process, usuay performed in different ocations. Figure 2 shows the resuts of an LCA comparison recenty conducted by Ricardo. Figure 2: Summary of embedded GHG emissions for ight-duty vehices Source: Ricardo review of pubished LCA studies (2018) 20 Concawe Review Voume 27 Number 1 Juy 2018

5 In terms of the BEV emissions, the LCA impact of the battery is mainy caused by the production chains of three components: the battery ces, the cathode and the anode, comprising together approximatey 55% to 85% of the battery s tota impact. [6] Tabe 3 summarises the contribution of different eements to the fina embedded CO 2 emissions associated with the battery manufacturing (and recycing) process. Tabe 3: Summary of the embedded CO 2 emissions in the battery manufacturing (and recycing) process kg CO 2 -e/ kwh battery Component Raw materia and refining a Battery grade materia production (incuding mining and refining) b Manufacturing (component and ce + battery assemby) Recycing Anode Cathode (90 c ) Eectroyte Separator <0.5 Approx. 1 Ce case <0.1 Approx. 1 Battery case Cooing Battery management system (est.) < TOTAL (216) Pyro: 15 Hydro: -12 Most ikey vaue (based on the assessment of transparency and scientific method done in the report) a Exampe based on materia needed for a 253-kg battery. (Ref. Eingsen et a. (2014), [6] and data from Tabe 14 of the IVL report [4] where the varied resuts from using different crade to gate datasets for materia extraction and production are iustrated.) b Ranges based on a review of battery LCAs. (Ref. Tabe 15 in the IVL report. [4] ) c Vaues in brackets are based on a report with approximate assumptions regarding processing materias. [7] Source: IVL (2017) [4] The eectricity mix used during the manufacturing process of the different battery components has a significant impact and, as iustrated in Figure 3 on page 22, notabe differences can be observed depending on the ocation (country) where both the manufacturing and assemby processes take pace. As extreme cases, and as extreme references for individua countries, NTNU estimates that the potentia impact of moving from a coa-based eectricity mix to a purey hydro-based country can be up to 4 t CO 2 -e, when an NMC ithium ion battery is considered (253 kg weight and 26.6 kwh energy capacity). Concawe Review Voume 27 Number 1 Juy

6 Figure 3: Sensitivity anaysis with respect to the source of eectricity for battery ce manufacture (resuts incude production and manufacturing). Impact category: goba warming potentia (2013 data) Source. Eingsen et a. (2014) [6] Impact of fue consumption (ICEs) and eectricity mix (BEVs) Life-cyce phases definition of terms When an LCA is appied to a vehice, it incudes the CO 2 emissions from manufacturing and disposing of the vehice itsef, as we as the CO 2 emissions from producing and suppying the fue to the vehice and consuming the fue in the vehice. The CO 2 emissions from producing and suppying the fue to the vehice are referred to as we-to-tank (WTT) emissions, whie the CO 2 emissions from consuming the fue in the vehice are referred to as tankto-whees (TTW) emissions. For BEVs, when ony the TTW CO 2 emissions are accounted, the BEV is considered as a zero-co 2 vehice due to the absence of tai pipe emissions. However, the WTT approach brings additiona CO 2 emissions into the whoe picture, i.e. the emissions associated with the production of the eectricity consumed and the energy osses from the eectricity generation site to the recharging device. Besides this definition, there is currenty an ongoing debate addressing how to consider the additiona TTW-reated osses associated with the battery recharging process, incuding the use of externa charging devices. Eectricity mix Currenty, Europe has a wide range of eectricity power generation technoogies across different countries ranging from coa based nationa eectricity mixes to mainy renewabe ones. In this context, the EU eectricity mix concept incuded in the study is considered as a reference point, whie individua assessments at country eve need to be conducted to produce specific scenarios that can be used to inform different stakehoders, incuding end users, about the LCA CO 2 performance of different aternatives. 22 Concawe Review Voume 27 Number 1 Juy 2018

7 Based on the assumptions mentioned in Tabe 2 (page 19), Concawe has conducted LCA-based comparisons of the CO 2 emissions of a compact cass (C-segment) vehice, for both BEV and ICE, where the impact of the nationa eectricity mix on the use phase is expored (see Figure 4). Figure 4: The impact of nationa eectricity mix on the use phase of two C-segment vehices: Nissan Leaf (BEV) vs Nissan Pusar (diese ICE) 1 battery/150,000 km NB It shoud be noted that the figure considers that the manufacturing emissions wi be the same regardess of the country in which the BEVs or ICEs are driven. Tabe 4: Specific inputs : C-segment CONSUMPTION AND WEIGHT* BATTERY SIZE AND RANGE CO 2 INTENSITY OF ELECTRICITY MIX Nissan Leaf (BEV), 109 hp 150 Wh/km (without osses) Kerb weight: 1,570 kg 30 kwh battery 250 km driving range EU mix (350 g CO 2 /kwh) Higher range: Poand mix (750 g CO 2 /kwh) Lower range: Sweden mix (20 g CO 2 /kwh) Nissan Pusar (diese ICE), 110 hp 3.8 /100 km (NEDC) (Kerb weight: 1,352 kg) * Data from Nissan brochures This comparison shows that the tota amount of CO 2 emitted during the ifetime of a Nissan Pusar (diese) can be simiar to the Nissan Leaf (BEV) when the eectricity mix incudes a arge fraction of coa powered generating pants, as in Poand. When the Nissan Leaf is compared to the C-segment benchmark data pubished by the European Commission, the ifetime CO 2 emissions for the Nissan Leaf are greater than for the diese vehice for the stated eectricity mix. Concawe Review Voume 27 Number 1 Juy

8 Athough Figure 4 shows the contribution of each stage to the tota CO 2 emissions, the evoution of the emissions aong the driven distance during the whoe ifetime of the vehice eads to interesting concusions (see Figure 5). Figure 5: The impact of nationa eectricity mix per distance driven for two C-segment vehices: Nissan Leaf (BEV) vs Nissan Pusar (diese ICE) At the beginning of the ife of the vehices, the BEVs have embedded emissions that are doube those of the equivaent ICE powertrains, due to the battery manufacturing process. Once in use, and during the first 50,000 km driven, the emissions from the diese fueed vehice (based on the C-segment benchmark vehice) woud remain ower than the overa emissions from a BEV when the anticipated 2030 EU average eectricity mix is considered. The CO 2 emissions for the diese vehice can remain ower than for the Nissan Leaf from 30,000 to greater than 150,000 km, depending on the eectricity mix used. Over the fu ife cyce of the vehice, the emissions from the use of an eectric vehice are eventuay ower than those from an equivaent ICE powertrain vehice, except in countries with a high reiance on coa. However, it is cear that on such an LCA basis, there are CO 2 emissions from the production and use of eectric vehices which shoud be taken into account in any assessment of the potentia for eectric vehices to contribute to goba GHG emission targets. 24 Concawe Review Voume 27 Number 1 Juy 2018

9 Size of the vehice The LCA methodoogy aows the comparison of different powertrains across different vehice segments, from a sma B-cass vehice to a arger D-cass or a uxury one. In 2016, NTNU presented the concept of the fossi enveope [8] which shows that the tota CO 2 emissions are heaviy dependent on the size of the vehice/battery chosen. Therefore, a comparison between individua vehices beonging to the same segment is crucia to conduct a comprehensive LCA. a) Sma vehices (B-segment) The category of subcompact vehices comprises a wide range of vehices with power and weights simiar to some of those considered as compact vehices. In this anaysis, a Mercedes Benz B-Cass and BMW i3 were initiay chosen as representatives of this BEVs B cassification. However, the weight of the Mercedes Benz B-Cass (1,700 kg) was more simiar to a C cassification vehice and, therefore, the comparison is focused on the BMW i3 (1,300 kg) with a 33 kw battery package. A these vehices have higher power than equivaent ICE B-segment vehices where consumers may opt for better fue efficiency (smaer size) in ess powerfu vehices than in other segments. Actuay, this customer choice between more powerfu vs more efficient vehices is a constant across a the passenger car casses but it is especiay important in the B-segment where higher fue efficiencies can be achieved. Figure 6 compares the BMW i3 (one of the smaer BEVs in the market) with the Mercedes B-Cass and a Renaut Cio IV B-segment ICE vehices. When the same approach described in Figure 5 is appied to these subcompact vehices, the break-even points for a BEV are between 20,000 and 60,000 km, usuay sooner than for the C-segment vehice but depending heaviy on the eectricity mix and the power of the ICE vehice chosen. In cases where a ower-hp ICE B-segment vehice is chosen (e.g. the Renaut Cio IV), the ife-cyce emissions are ess than those of a BEV with a Poish eectricity mix. Figure 6: The impact of nationa eectricity mix per distance driven for three B-segment vehices: BMW i3 (BEV) vs Mercedes Benz B-Cass (diese ICE) vs Renaut Cio IV GrandTour (diese ICE) Concawe Review Voume 27 Number 1 Juy

10 Tabe 5: Specific inputs : B-segment CONSUMPTION AND WEIGHT* BATTERY SIZE AND RANGE CO 2 INTENSITY OF ELECTRICITY MIX BMW i3 (BEV) 161 Wh/km (without osses) Kerb weight: 1,300 kg 33 kwh battery 180 km driving range EU mix (350 g CO 2 /kwh) Higher range: Poand mix (750 g CO 2 /kwh) Lower range: Sweden mix (20 g CO 2 /kwh) Renaut Cio (diese ICE) 3.2 /100 km (NEDC) (Kerb weight: 1,200 kg) Mercedes Benz B-180 (diese ICE), 109 hp 3.6 /100 km (NEDC) (Kerb weight: 1,395 kg) * Data from Mercedes brochures b) Large vehices (D-segment/best-in-cass) In the case of arge vehices, the embedded emissions associated with the manufacturing stage increase significanty due to the combination of the arger sizes of both of the vehice and the battery used to increase the driving range (representing 45% of the tota CO 2 emissions in the seected BEV exampe). The ocation of the battery manufacturing and assembing faciities has a arge impact on ifetime emissions in this vehice segment. Aso, due to the higher embedded CO 2 emissions for these D-segment BEVs, and as the vehices in this segment are typicay used to drive onger distances on more frequent journeys, ensuring that no battery repacement woud be required aong their ifetimes becomes the key factor in the comparison versus an equivaent ICE vehice. Assuming that ony one battery is used, a BEV consuming an eectricity mix cose to the EU mix woud need to be driven more than 100,000 km to reach the crossover point at which both powertrains reach parity in CO 2. When a Poish eectricity mix is used, the anaysis shows that an ICE diese vehice emits ess CO 2 than a BEV during its whoe ifetime. Figure 7: The impact of nationa eectricity mix and battery repacement on the use phase of two D-segment vehices: Tesa Mode S (BEV) vs Audi A4 (diese ICE) 150,000 km Source: Concawe anaysis based on data from Tesa, Audi, NTNU, T&E and IVL 26 Concawe Review Voume 27 Number 1 Juy 2018

11 Figure 8: The impact of nationa eectricity mix per distance driven for two D-segment vehices: Tesa Mode S (BEV) vs Audi A4 (diese ICE) Tabe 6: Specific inputs: D-segment CONSUMPTION AND WEIGHT* BATTERY SIZE AND RANGE CO 2 INTENSITY OF ELECTRICITY MIX Tesa Mode S (BEV) 181 Wh/km (without osses) Kerb weight: 2,100 kg 100 kwh battery > 250 km driving range (NTNU vaue) EU mix (350 g CO 2 /kwh) Higher range: Poand mix (750 g CO 2 /kwh) Lower range: Sweden mix (20 g CO 2 /kwh) Audi A7 3.0 TDI (diese ICE) 4.7 /100 km (NEDC) (Kerb weight: 1,800 kg) * Data from Tesa and Audi brochures, bogs and NTNU data c) The eectricity enveope Figure 9 on page 28 shows the minimum distance that a vehice woud need to be driven to reach CO 2 emission parity between a BEV and an ICE powertrain for different eectricity mixes, simiar to the fossi fue enveope deveoped by NTNU for ICE vehices. The concept of an eectricity mix enveope can be appied to expore the importance of the carbon intensity of the eectricity mix consumed by BEVs. This type of figure at a nationa eve coud hep to inform different stakehoders, incuding consumers and poicymakers, of the best avaiabe options, considering vehice cass, battery manufacturing ocations and expected distance driven. Concawe Review Voume 27 Number 1 Juy

12 Figure 9: Exampe of an eectricity mix enveope based on seected vehices Source: Concawe Fue consumption: the roe of ow-carbon iquid fues Concawe is aso exporing how future high efficiency interna combustion engine technoogies, combined with ow-carbon fues, have the potentia to deiver significant CO 2 savings. To guide future research and poicy, the same LCA anayses can be deveoped to assess the mitigation potentia of different pathways. As shown by the green area in Figure 10, the combination of ICE and hybridisation has the potentia to provide ife-cyce CO 2 emission savings comparabe with forecasted figures for future improved BEVs powered by eectricity generated mainy using renewabe sources (red ine). Ricardo is currenty conducting an anaysis assessing the LCA emissions associated with ow-carbon fues combusted in the most efficient interna combustion/hybrid vehices. [1] The preiminary resuts confirm that, by 2050 for certain advanced biofues and power-to-iquid technoogies, the combination of highy efficient ICE powertrains and ower-carbon fues are ikey to give simiar reductions in ife-cyce CO 2 emissions when compared with BEV vehices powered by a highy decarbonized eectricity mix. Figure 10: The potentia roe of ow-carbon fues in an LCA (conceptua approach) Source: Concawe interna work 28 Concawe Review Voume 27 Number 1 Juy 2018

13 Concusions This is the first pubished artice associated with Concawe s ongoing research on ife-cyce anaysis and the potentia roe of ow-carbon fues that is being undertaken as part of the ong-term strategy of the refining industry. [9,10] Severa concusions can be drawn from the resuts presented: Methodoogy: LCA is a scientificay sound, we accepted methodoogy aowing the comparison of different powertrains on the same basis. Currenty, there is a need for more data from manufacturers, especiay regarding batteries, to improve the accuracy of the LCAs conducted in the transport sector, and a recognised standardised basis for conducting LCAs. Comparisons shoud be performed at the country eve and by vehice cassification to provide meaningfu resuts. Resuts: When comparing eectric vehices with conventiona powertrains, the reative contribution to the tota ife-cyce emissions associated with the different phases of energy or fue generation, materias extraction, manufacturing and production a need to be incuded. The energy mix used during the battery manufacturing process has a significant impact on the tota CO 2 emitted during the ife of the vehice. The carbon intensity of the eectricity mix used to recharge the vehices has a strong infuence on the ife-cyce emissions of eectric vehices. The break-even distance for the ife-cyce CO 2 emissions of eectric vehices compared to conventiona vehices is dependent on the vehice size as we as the eectricity mix. B-segment vehices have the shortest break-even distance, whie D-segment vehices have the ongest breakeven distance. For an eectricity mix that is heaviy dependent on coa and heavier vehices, the ife-cyce CO 2 emissions of eectric vehices wi be greater than for conventiona vehices. Future high-efficiency ICE technoogies, combined with ow-carbon fues, have the potentia to deiver significant CO 2 savings across a segments, simiar to BEVs using a argey renewabe-based eectricity mix. Finay, the vehices presented in this study are exampes chosen to iustrate the main concepts of the LCA methodoogy, stressing the importance of different parameters in the fina CO 2 emissions associated with each powertrain considered. Concawe is wiing to engage with externa stakehoders to assist in deveoping a standard LCA methodoogy specificay for anaysing future fues and powertrains. Concawe Review Voume 27 Number 1 Juy

14 References 1. Preiminary concusions from Ricardo s recent anaysis, commissioned by Concawe, assessing the ifecyce CO 2 impact of both a mass-bev scenario and a ow-carbon fue scenario (report to be pubished in Q4 2018). 2. Moro, A. and Lonza, L. (2017). Eectricity carbon intensity in European Member States: Impacts on GHG emissions of eectric vehices. In press. 3. NTNU (2017). Life cyce assessment of eectric vehices. Norwegian University of Science and Technoogy IVL (2017). The Life Cyce Energy Consumption and Greenhouse Gas Emissions from Lithium-Ion Batteries. Mia Romare and Lisbeth Dahöf, IVL, Sweden. 5. BMVI (2016). Bewertung der Praxistaugichkeit und Umwetwirkungen von Eektrofahrzeugen. (Assessment of the feasibiity and environmenta impacts of eectric vehices.) Federa Ministry of Transport and Digita Infrastructure, Germany. 6. Eingsen et a. (2014). Life Cyce Assessment of a Lithium-Ion Battery Vehice Pack. In Journa of Industria Ecoogy, Vo. 18, No. 1, pp Majeau-Bettez, G. et a. (2011). Life Cyce Environmenta Assessment of Lithium-Ion and Nicke Meta Hydride Batteries for Pug-In Hybrid and Battery Eectric Vehices. In Environmenta Science & Technoogy, Vo. 45, Issue 10, pp (See aso addition/correction in Vo. 45, Issue 12, p ) 8 Eingsen et a. (2016). The size and range effect: ifecyce greenhouse gas emissions of eectric vehices. In Environmenta Research Letters, Vo. 11, No Vision The Low Carbon Pathways Project. Pan_The-Low-Carbon-Pathways-Project.pdf 30 Concawe Review Voume 27 Number 1 Juy 2018