Reducing CO2 Emissions from Road Transport in the European Union: an Evaluation of Policy Options

WORKING PAPER 2016-10

Reducing CO2 emissions from road transport in the European Union: An evaluation of policy options

Author: Joshua Miller Date: 16 May 2016

Keywords: European Union, light-duty vehicles, heavy-duty vehicles, emissions trends, CO2 standards, fiscal policies, GHG emissions

1. Introduction include medium trucks, heavy trucks, 2. Methods and buses. Together, LDVs and HDVs Adopted in October 2014, the account for 95% of GHG emissions European Union’s 2030 climate POLICY OPTIONS FOR LOW- from non-ETS transport3 and one-fifth and energy framework established CARBON ROAD TRANSPORT of total EU GHG emissions, and CO a binding target for GHG emissions 2 accounts for 99% of GHG emissions We developed one or more policy of 40% below 1990 levels by 2030.1 from road transport. options within each of six categories: It has two components. Sectors inclusion of road transport in the EU covered by the EU emission trading Specifically, the paper: ETS, LDV efficiency, HDV efficiency, system (ETS) must reduce emissions electric-drive vehicles, biofuels, and to 43% below a 2005 baseline by • Separately summarizes the fuel taxation. We then ordered and

2030. Non-ETS sectors, which include impacts of CO2 standards for sequentially combined these policy transport, buildings, and agriculture, LDVs and HDVs options to construct the “narratives” must reduce emissions to 30% below • Compares the marginal benefits described in the following section. a 2005 baseline in 2030. Of the of moderate and stringent CO2 non-ETS sectors, transport (including standards, including a policy Inclusion of road transport in road, rail and inland waterways) is the option to reduce the gap between the EU ETS largest contributor of GHG emissions. laboratory and real-world fuel One proposal for addressing road Radically reducing emissions from consumption transport is essential to meeting the transport GHG emissions has been 2 • Evaluates direct CO emissions, to include the sector in the EU ETS, 2030 climate goals. 2 which are the focus of the 2030 effectively putting a price on GHG This paper describes and evaluates climate target, as well as indirect emissions from the combustion of potential low-carbon road transport emissions impacts of electric fuels for road transport.4 Based on an policies for achieving the 2030 target vehicles and biofuels for non-ETS sectors. It focuses on CO 2 • Constructs sequential 4 ExxonMobil Petroleum & Chemical BVBA from light-duty vehicles (LDV), which “narratives” (policy pathways) (2015). “Perspectives on the future of include passenger cars and vans, and that allow comparison of the European transport.” Retrieved from http:// heavy-duty vehicles (HDV), which exxonmobil.com/Europe-English/Files/ marginal benefits of individual EU_Transportation.pdf; Martin, I. (2013). policies as well as the impacts of “RE: DG Climate Action consultation on 1 European Commission (2016). “2030 multiple policies combined. the report from the Commission to the climate & energy framework.” Accessed European Parliament and the Council – The 4 Apr 2016 at http://ec.europa.eu/clima/ state of the European carbon market in policies/strategies/2030/index_en.htm 2012.” Royal Dutch Shell Plc. Retrieved from 2 Transport & Environment (2016). How http://ec.europa.eu/clima/consultations/ the European car industry plans to meet docs/0017/organisations/shell_en.pdf; the climate challenge. Retrieved from Achtnicht, M., von Graevenitz, K., Koesler, http://www.transportenvironment.org/ S., Löschel, A., Schoeman, B., Angel Tovar publications/how-european-car-industry- 3 Domestic aviation emissions are covered by Reanos, M. (2015) “Including road transport plans-meet-climate-challenge the EU ETS. in the EU-ETS – An alternative for the

Table 1. Adjustment from NEDC to WLTP targets for passenger cars

Test cycle for Current policies Moderate targets Stringent targets NTE vehicle certification Variable 2014 2021 2025 2030 2025 2030 2025 2030 Target [g/km] 123 95 78 60 68 42 NEDC Real-world gap [%] 38% 49%

Real-world CO2 [g/km] 170 142 Target [g/km] 109 90 69 78 48 78 48 WLTP Real-world gap [%] 23% 31% 31% 31% 31% 15% 15%

Real-world CO2 [g/km] 134 118 90 102 63 90 55 Adjustment from NEDC to WLTP targets 115% Bold indicates nominal targets for the active certification test cycle.

ETS certificate price of €5 per ton of Moderate targets switch to WLTP in 2017/18 decreases In April 2013, the European Parliament CO2, including road transport in the the real-world fuel consumption gap ETS would translate to an effective asked the European Commission to from 49% under the NEDC (with fuel tax of 1 Euro cent per liter.5 provide by the end of 2016 an impact a 2021 target of 95 g/km) to 23%

assessment of LDV CO2 standards under the WLTP (with a 2021 target Light-duty vehicle efficiency with an indicative range of 68-78 of 109 g/km), thereby improving g/km in 2025, as measured by the real-world emissions from 142 g/ We developed four policy options NEDC.8 Assuming a conversion factor km (under the NEDC) to 134 g/km for CO standards that promote the 2 of 1.15 from NEDC to WLTP,9 the (under the WLTP). The nominal and efficiency of new LDVs. ‘Moderate targets’ scenario assumes real-world targets for the ‘Moderate Current policies a WLTP target of 90 g/km for cars in targets’, ‘Stringent targets’, and ‘NTE’ Under adopted EU standards, new 2025 (78 g/km on NEDC), tightening scenarios are shown in Table 1.11 cars and vans must meet targets to 69 g/km in 2030 (60 g/km on Stringent targets of 95 g/km in 2021 and 147 g/km NEDC), and equivalent percent 10 The ‘Stringent targets’ scenario in 2020, respectively, as measured reductions for vans. The ‘Moderate assesses the impacts of a 78 g/km on the NEDC test cycle. The ICCT targets’ scenario assumes that a WLTP target for cars in 2025 (60 has documented an increasing gap g/km on NEDC), tightening to 48 between these official CO values and 2 8 Mock, Peter (2013, May 5). “EU vote on g/km in 2030 (42 g/km on NEDC) real-world performance, from 8% in cars CO2: 95 g/km in 2020, 68-78 g/km (Table 1).12 As in the ‘Moderate 2001 to 38% in 2014.6 Under current in 2025.” Washington DC: International targets’ scenario, similar percentage policies, this gap could increase to Council on Clean Transportation. Retrieved from http://www.theicct.org/blogs/staff/ reductions are assumed for vans. 49% in 2020 if the NEDC test cycle eu-vote-cars-co2 were to continue being used for 9 According to ICCT’s assessment, the Stringent targets and not-to- official CO certification.7 adjustment factor from an NEDC to a WLTP exceed limit 2 target from a technical point of view should not be higher than 1.08. At the same time, The ‘High ambition plus NTE’ scenario it is likely that the adjustment factor from evaluates the impact of adding a not- the European Commission’s NEDC-WLTP to-exceed (NTE) limit of 15% starting future?” ZEW. Retrieved from http:// correlation exercise will be closer to 1.15, ftp.zew.de/pub/zew-docs/gutachten/ which is why a factor of 1.15 is assumed in RoadTransport-EU-ETS_ZEW2015.pdf this analysis. Source: Mock, P., Kühlwein, 5 Mock, P., Tietge, U., German, J., and J., Tietge, U., Franco, V., Bandivadekar, A., 11 Tietge, U., Zacharof, N., Mock, P., Franco, Bandivadekar, A. (2014). “Road transport and German, J. (2014). “The WLTP: How a V., German, J., Bandivadekar, A., Ligterink, in the EU Emissions Trading System: An new test procedure for cars will affect fuel N., and Lambrecht, U. (2015). “From engineering perspective.” Washington consumption values in the EU.” Washington Laboratory to Road: A 2015 Update of DC: International Council on Clean DC: International Council on Clean Official and “Real-World” Fuel Consumption Transportation. Retrieved from http://www. Transportation. Retrieved from http://www. and CO2 Values for Passenger Cars.” ICCT, theicct.org/sites/default/files/publications/ theicct.org/wltp-how-new-test-procedure- TNO and IFEU. Retrieved from http://www. ICCT_EU-ETS-perspective_20141204.pdf cars-will-affect-fuel-consumption-values-eu theicct.org/laboratory-road-2015-update 6 Tietge, U., Zacharof, N., Mock, P., Franco, 10 Hill, Nikolas (2016). “SULTAN modelling 12 Hill, Nikolas (2016). “SULTAN modelling V., German, J., Bandivadekar, A., Ligterink, to explore the wider potential impacts of to explore the wider potential impacts of N., and Lambrecht, U. (2015). “From transport GHG reduction policies in 2030.” transport GHG reduction policies in 2030.” Laboratory to Road: A 2015 Update of Prepared by Ricardo Energy & Environment Prepared by Ricardo Energy & Environment Official and “Real-World” Fuel Consumption for the European Climate Foundation. for the European Climate Foundation. and CO2 Values for Passenger Cars.” ICCT, Retrieved from http://europeanclimate. Retrieved from http://europeanclimate. TNO and IFEU. Retrieved from http://www. org/wp-content/uploads/2016/02/ECF- org/wp-content/uploads/2016/02/ECF- theicct.org/laboratory-road-2015-update Transport-GHG-reduction-for-2030_Final_ Transport-GHG-reduction-for-2030_Final_ 7 Ibid. Issue21.pdf Issue21.pdf

2 INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION WORKING PAPER 2016-10 REDUCING CO2 EMISSIONS FROM ROAD TRANSPORT IN THE EUROPEAN UNION: AN EVALUATION OF POLICY OPTIONS

in 2025. Such a limit would set a about 1%, including BEVs, PHEVs, battery electric, and 1.2% hydrogen maximum legal difference between and FCEVs. In the current policies fuel cell vehicles).17

CO2 emissions under reasonable scenario, the EV market share is con- real-world driving conditions and servatively assumed to remain at 1% Biofuels laboratory tests under the WLTP. in 2030. This conservative baseline Current policies Compliance with the NTE limit would assumption is intended to give credit Under current policies, we assume be evaluated through confirmatory to policies in the e-drive transition first generation (1G) biofuels will testing of in-production vehicles. scenario for increased EV uptake reach 7% blending in 2020 and rather than assume that EV sales will remain constant thereafter.18 We increase regardless. In addition, this Heavy-duty vehicle efficiency assume the expansion of ethanol and assumption acknowledges that much 2020 introduction biodiesel from current blend levels of the EV uptake to date is driven by The EU is expected to require will be proportional to their current fiscal and non-fiscal incentives, and mandatory CO certification of HDVs blend levels. Given current blend 2 without these supporting policies, starting in 2018. Considering this levels of 3.4% for ethanol in petrol there is no guarantee that EVs will certification start date, HDV CO and 5.3% for biodiesel in diesel, we 2 capture a substantial share of the standards could be introduced as assume 5.2% ethanol in petrol and LDV market. early as 2020. Under this scenario, 8% biodiesel in diesel to reach 7% HDV CO standards are assumed to blending overall.19 We note that 2 E-drive transition require 3% annual improvements This scenario assesses the impacts of 8% blending of biodiesel in diesel 20 from 2020 to 2030. This assumption a decades-long transition to electrify exceeds the current EU blend limit translates to a 26% reduction in CO2 the LDV fleet by strengthening and but that this may occur as a result of from new HDVs over the period 2020 expanding international best practice use of higher biodiesel blends such 13 21 to 2030. policies to spur electric-drive (e-drive) as B20 or B100. We assume a split vehicle deployment. Norway and the of 45% starch ethanol and 55% sugar 2025 introduction Netherlands have already adopted ethanol for the 1G ethanol category Alternatively, if HDV CO standards 2 based on a 2016 projection of 14 several of these policies and are now were introduced starting in 2025, 22 seeing “electric vehicle deployment feedstock use. Second generation such standards might require 3% that is more than ten times the (2G) biofuels reach 0.5% in 2020 annual improvements from 2025 to international average” as identified and remain constant thereafter. We 2030. This assumption translates to a in a survey of efforts to establish a assume this is 75% cellulosic ethanol 14% reduction in CO from new HDVs 2 global market for electric vehicles and 25% cellulosic/waste-oil diesel in 2030 compared to a 2025 baseline. (EVs).15 Policy actions in this scenario include consumer incentives, support 17 Ibid. Electric-drive vehicles for charging infrastructure, and 18 Hill, Nikolas (2016). “SULTAN modelling Current policies continued support for R&D, driven by to explore the wider potential impacts of transport GHG reduction policies in 2030.” While some markets have substan- 16 aggressive EV deployment goals. Prepared by Ricardo Energy & Environment tially higher EV market share (Norway Consistent with the results of this for the European Climate Foundation. had ~20% and the Netherlands global survey, this analysis assumes Retrieved from http://europeanclimate. had ~6% in 2015), most European org/wp-content/uploads/2016/02/ECF- e-drive vehicles account for 23% of Transport-GHG-reduction-for-2030_Final_ markets had EV sales shares of new LDV sales in the EU by 2030 Issue21.pdf less than 1% in 2015. On average, (including 8.8% plug-in hybrids, 12.8% 19 Ibid. the EU had an EV market share of 20 Kampman, B. (CE Delft), Verbeek, R. (TNO), van Grinsven, A. (CE Delft), van Mensch, P. (TNO), Croezen, H. (CE Delft), and Patuleia, 15 Lutsey, Nic (2015). “Transition to a global A. (TNO) (2013). “Bringing biofuels on the 13 The scenarios for HDV standards consider zero-emission vehicle fleet: A collaborative market: Options to increase EU biofuels only improvements to new HDV efficiency agenda for governments.” Washington volumes beyond the current blending rather than in-use improvements. In-use DC: International Council on Clean limits.” Retrieved from https://ec.europa.eu/ improvements (such as ITS, green freight, Transportation. Retrieved from http://www. energy/sites/ener/files/documents/2013_11_ and driver training) could potentially result theicct.org/transition-global-zero-emission- bringing_biofuels_on_the_market.pdf in additional benefits; however, these in-use vehicle-fleet-collaborative-agenda- 21 Higher biodiesel blends could take improvements are less likely than new governments longer to materialize due to vehicle and vehicle standards to have fleetwide emission 16 While not all of these actions have been infrastructure requirements. benefits (allowing time for vehicle turnover). undertaken in each leading EV market, 22 Flach, B., Lieberz, S., Rondon, M., 14 While some might consider a later these actions have been identified as Williams, B., and Teiken, C. (2015). “EU implementation date to be more realistic, international best practice policies to Biofuels Annual 2015.” GAIN Report HDV standards would have to be promote EVs. Each market should choose Number: NL5028. Retrieved from http:// implemented at least several years before the policies that are best suited to its gain.fas.usda.gov/Recent%20GAIN%20 2030 to have a meaningful impact toward socioeconomic, political and geographic Publications/Biofuels%20Annual_The%20 the 2030 climate target. characteristics. Hague_EU-28_7-15-2015.pdf

WORKING PAPER 2016-10 INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION 3 REDUCING CO2 EMISSIONS FROM ROAD TRANSPORT IN THE EUROPEAN UNION: AN EVALUATION OF POLICY OPTIONS

Table 2. Assumed biofuel shares of petrol and diesel-equivalent fuels

Conventional Scenario fuel Biofuel 2015 2020 2025 2030 Starch Ethanol 1.5% 2.3% 2.3% 2.3% Petrol Sugar Ethanol 1.9% 2.9% 2.9% 2.9% Current policies Cellulosic Ethanol 0.0% 1.0% 1.0% 1.0% Vegetable oil-based Biodiesel 5.2% 8.0% 8.0% 8.0% Diesel Cellulosic/Waste-oil Diesel substitute 1.0% 1.2% 1.2% 1.2% Starch Ethanol 1.5% 1.5% 1.5% 1.5% Petrol Sugar Ethanol 1.9% 1.9% 1.9% 1.9% New fuels Cellulosic Ethanol 0.0% 1.0% 4.4% 7.7% Vegetable oil-based Biodiesel 5.2% 5.2% 5.2% 5.2% Diesel Cellulosic/Waste-oil Diesel substitute 1.0% 1.2% 2.0% 2.7%

Table 3. Definition of low-carbon transport narratives

Policy category Narrative ETS LDV EFFICIENCY HDV EFFICIENCY E-DRIVE BIOFUELS FUEL TAXES 1. Baseline Not included Current policies None Current policies Current policies Current taxes 2a. ETS only ETS Current policies None Current policies Current policies Current taxes 2b. Moderate LDV Moderate Not included None Current policies Current policies Current taxes efficiency targets 2c. High LDV Stringent Not included None Current policies Current policies Current taxes efficiency targets 2d. High LDV Stringent Not included None Current policies Current policies Current taxes efficiency NTE targets and NTE 3a. Moderate HDV Stringent targets 2025 Not included Current policies Current policies Current taxes efficiency and NTE introduction 3b. High HDV Stringent targets 2020 Not included Current policies Current policies Current taxes efficiency and NTE introduction Stringent targets 2020 E-drive 4. Electrification Not included Current policies Current taxes and NTE introduction transition Stringent targets 2020 E-drive 5. Biofuels Not included New fuels Current taxes and NTE introduction transition Stringent targets 2020 E-drive 6. Fuel taxation Not included New fuels New taxes and NTE introduction transition Italics indicate no change from previous narrative. Bold indicates change from previous narrative. Narrative fill colors will appear in figure legends. substitute. In addition, we assume fuels, we assume a linear increase in Fuel taxes 2 million tonnes of oil equivalent blend levels from 0.5% in 2020 to 4% Current taxes (Mtoe) biodiesel from waste oils is in 2030. As above, we assume that Current fuel taxes in the EU vary produced in each year. 75% of 2G fuels are cellulosic ethanol substantially across member states, and 25% are Fischer-Tropsch (F-T) from 0.36 to 0.75 Euros per liter for New fuels renewable diesel; and in addition, petrol and from 0.33 to 0.67 Euros In the ‘New fuels’ scenario, we that 2 Mtoe biodiesel from waste per liter for diesel fuel.24 These fuel assume 1G biofuels are capped at oils is produced in each year. Table taxes are reflected in the current and 2015 levels (3.4% ethanol in petrol; 2 summarizes the assumed biofuel 23 projected baseline. 5.2% biodiesel in diesel). For 2G shares of petrol and diesel-equivalent fuels for each biofuels scenario.

23 Hill, Nikolas (2016). “SULTAN modelling 24 European Environment Agency (2016). to explore the wider potential impacts of Retrieved from http://europeanclimate. “Fuel prices.” Accessed 4 Apr 2016 at transport GHG reduction policies in 2030.” org/wp-content/uploads/2016/02/ECF- http://www.eea.europa.eu/data-and- Prepared by Ricardo Energy & Environment Transport-GHG-reduction-for-2030_Final_ maps/indicators/fuel-prices-and-taxes/ for the European Climate Foundation. Issue21.pdf assessment-5

4 INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION WORKING PAPER 2016-10 REDUCING CO2 EMISSIONS FROM ROAD TRANSPORT IN THE EUROPEAN UNION: AN EVALUATION OF POLICY OPTIONS

New taxes 1,200 In the absence of a specific proposal to strengthen fuel taxes across the EU, 1,000 the ‘new taxes’ scenario assesses the 1. Baseline potential impact of all member states

] 2a. ETS only increasing fuel taxes by an average of 800 2b. Moderate LDV e ciency 20 Euro cents per liter for petrol and year diesel fuel. Such a tax is assumed to 2c. High LDV e ciency [Mt/ phase in gradually, starting at 5 Euro 2 600 2d. High LDV e ciency NTE CO cents in 2020, increasing to 10 Euro 3a. Moderate HDV e ciency cents in 2025, and reaching 20 Euro 3b. High HDV e ciency rect 400

cents in 2030. Di 4. Electriﬁcation 5. Biofuels 200 NARRATIVES 6. Fuel taxation For the purpose of evaluating the numerous possible policy com- 0 2005 2010 2015 2020 2025 2030 binations, we’ve constructed ten Figure 1. Direct CO from LD and HD vehicles by narrative, 2005-2030. Numbered alternate “narratives” illustrating 2 narratives are additive, such that narrative ‘6’ includes ‘Fuel taxation’, ‘Biofuels’, different possible policy trajectories ‘Electrification’, ‘High HDV efficiency’, and ‘High LDV efficiency NTE.’ (Table 3). Lettered narratives with the same number (e.g. ‘2a’ and ‘2b’) In the baseline narrative, direct reduce emissions by 93 Mt compared represent mutually exclusive futures: emissions are projected to remain to the ‘ETS only’ narrative, and 95 for example, the ‘ETS only’ narrative relatively flat from 2015-2030, Mt25 compared to the 2030 baseline. More stringent LDV targets (‘2c’) includes the road transport sector with adopted LDV CO2 standards within the ETS but assumes no further largely offsetting projected growth could reduce direct emissions by an adoption of low-carbon transport in passenger and freight activity. additional 49 Mt in 2030 compared policies. Narratives are numbered Importantly, the numbered narratives to moderate targets (‘2b’), and sequentially in order of ease of imple- are additive: that is, if narrative ‘6. adding a not-to-exceed limit of 15% mentation (starting with ‘ETS only’, Fuel taxation’ were implemented to LDV standards (‘2d’) could further then ‘LDV efficiency’, etc.). Within (which includes all of the policies in augment these reductions by 25 Mt. each number, the narrative with the narratives ‘2d’, ‘3b’, ‘4’, and ‘5’, as well In total, ambitious LDV standards letter closest to ‘z’ represents the as new fuel taxes), direct emissions in that include a not-to-exceed limit most ambitious. could reduce 169 Mt compared to the 2030 could be reduced to 677 MtCO2 (hereafter abbreviated as Mt), or 24% baseline in 2030 (narrative ‘1’ minus As illustrated in Table 3, narratives below a 2005 baseline. ‘2d’), about 80% greater than the with higher numbers are additive, impact of moderate LDV CO2 targets meaning they include the most without such a limit. ambitious option (letter) of each DIRECT CO2 IMPACTS IN 2030 lower numbered narrative. Thus, Figure 2 compares emissions Introduction of HDV standards in narrative ‘6. Fuel taxation’ includes all estimates for each narrative in 2025 (‘3a’) could avoid 17 Mt in of the policies in narratives ‘2d’, ‘3b’, 2030 with the 2005 baseline. The 2030. Earlier introduction of HDV ‘4’, and ‘5’, as well as the policy for short, transparent bars indicate the standards in 2020 (‘3b’) could avoid ‘New taxes’. Further explanation of 55 Mt by 2030 – more than three marginal CO2 reduction from the narratives is provided in Appendix D. previous narrative, starting with times the benefits of introduction in a comparison of the ‘ETS only’ 2025. Assuming ambitious actions narrative (‘2a’) against the baseline have already been taken to promote 3. Results in 2030. Including road transport in LDV and HDV efficiency (‘2d’ and the EU ETS would have a very small ‘3b’), electrifying the LDV market DIRECT CO2 EMISSIONS FROM impact due to the low carbon price (‘4’) could avoid 19 Mt in 2030. 2005-2030 that is observed today (equivalent Expanding the deployment of 2G

Figure 2 compares direct CO2 to about 1 Euro cent per liter fuel). emissions from LDVs and HDVs by In contrast, the ‘Moderate LDV 25 Reported estimates are not rounded, efficiency’ narrative (‘2b’) could whereas figures show estimates rounded to policy narrative from 2005-2030. the nearest ton.

WORKING PAPER 2016-10 INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION 5 REDUCING CO2 EMISSIONS FROM ROAD TRANSPORT IN THE EUROPEAN UNION: AN EVALUATION OF POLICY OPTIONS

biofuels (‘5’) would only marginally 1,200 reduce direct emissions, since most 1,000 959 956 of the gains from 2G biofuels are 891 863

year] 814 790 773 800 736 716 716 classified as indirect emissions (see 677 [Mt/

Figure 4). Finally, strengthened fuel 2 600 taxes (‘6’) could reduce another 39 CO 400 MtCO2 in 2030, complementing the impacts of CO standards, electrifi- rect 2 200 93 Di 49 38 39 cation, and biofuels. 3 25 17 19 1 0 y y y y E io n ion t uels NT only at ienc ienc ienc ienc s eline DIRECT CO COMPARED TO A s eline of xa 2 S Bi iﬁc ta Ba Ba

2005 BASELINE ET

tr e c e c e c e c ien cy 1. 1. 5. uel Figure 3 reframes the absolute 2a . F e c E lec LD V LD V HD V HD V

emissions results shown in Figure 6. 4. te 2 in terms of a cumulative percent te LD V High High reduction from a 2005 baseline. . de ra de ra 2c 3b.

Under the ‘ETS only’ narrative, direct High Mo

Mo emissions would be an estimated 7.4% 2d. 2b. higher in 2030 than in 2005. 3a. 2005 2030

Figure 2. Direct CO2 from LD and HD vehicles by narrative, 2005 and 2030. Solid bars indicate direct emissions for each narrative. Transparent bars indicate marginal CO DIRECT AND INDIRECT CO2 2 IN 2030 reductions (reflecting the benefits of new policies in a given narrative), starting with a comparison against the 2030 baseline. Figure 4 summarizes the direct and indirect emissions impacts of 30% io n 24.0% each policy narrative in 2030. For ] 25% 19.5% 19.7% policies that focus on reducing petrol [%

du ct 20% 17.4%

re 13.2% and diesel consumption through 15% 11.3% t efficiency or pricing measures (added 10% 8.5% ase lin e 3.1% rcen in narratives numbered ‘2’, ‘3’ and b 5% -7.6% -7.4% ‘6’), the marginal indirect emissions pe 0%

ve -5% 2005 benefits tend to be proportional to ti -10% y E the direct benefits. For LDV elec- om ula fr ion io n m t uels NT

trification (‘4’), indirect emissions only at ien cy ienc ien cy ien cy s eline Cu of xa from electricity generation would S Bi iﬁc ta Ba

en cy

tr offset direct emissions benefits by 3 e c e c e c e c 1.

5. uel ec Mt in 2030. In contrast, for biofuels 2a. F El e ci

LDV LDV HDV HD V

(‘5’), considering indirect emissions 6. 4. te (including production emissions and te LD V High High

ILUC relative to conventional fuels) . 2c 3b. High

increases the net emission benefit by M ode ra M ode ra 26 Mt in 2030. These gains are attrib- 2d. 2b. 3a. utable to the fuel lifecycle benefits 2030 of second generation biofuels (in Figure 3. Cumulative percent reduction in direct CO2 from 2005 baseline by the ‘New fuels’ scenario) relative to narrative, 2030 first generation biofuels (which are projected to eclipse 2G shares under category as follows: decrease in business-as-usual current policies). transport GHG emissions.26 To • Baseline: Assuming no progress

beyond current policies, we 26 Hill, Nikolas (2016). “SULTAN modelling 4. Conclusions estimate a 7.6% increase in to explore the wider potential impacts direct CO from LDVs and HDVs of transport GHG reduction policies in This analysis yields a number of 2 2030.” Prepared by Ricardo Energy & over the period 2005 to 2030. findings that are relevant to near-term Environment for the European Climate In contrast, a recent Ricardo- Foundation. Retrieved from http:// decisions on EU climate policy. These AEA study estimated a 16% europeanclimate.org/wp-content/ findings are summarized by policy uploads/2016/02/ECF-Transport-GHG-

6 INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION WORKING PAPER 2016-10 REDUCING CO2 EMISSIONS FROM ROAD TRANSPORT IN THE EUROPEAN UNION: AN EVALUATION OF POLICY OPTIONS

the extent that regulators use ] 1,200 these studies to inform their year 1,000 959 956 891 863 design of low-carbon transport 814 790 773 [Mt/

800 736 716 716 policies, an increasing baseline 2 677 indicates a greater level of CO 600 policy action is needed to meet

ec t 352 400 342 341 310 the 2030 climate target for 294 285 279 266 269 243 230

non-ETS sectors. indir 200

• ETS inclusion: An ‘ETS only’ nd a 0 s y E

approach is entirely insufficient on ion ect ti t r uel NT only ienc ien cy ien cy ien cy

considering the magnitude of y s eline s eline of Di xa S Bi iﬁ ca ta Ba GHG reductions targeted and Ba

tr e c e c e c e c ienc 1. 1.

the minimal fuel price impact of 5. uel ec 2a. F El

including road transport in the e c

LD V LDV HD V HD V

6. 4. te

ETS. Even the ‘Fuel taxation’ te LD V High narrative, at roughly 20 times High

. 2c the stringency of the ‘ETS ode ra 3b. High

only’ approach in terms of the M ode ra 2d. 2b. assumed increase in fuel prices, 3a . M would impact baseline emissions 2005 2030 by less than 5% in 2030.27 Figure 4. Direct and indirect CO2 emissions from LD and HD vehicles by narrative, 2030. Solid bars indicate direct emissions, while grey filled bars indicate indirect emissions. • LDV CO2 standards: While moderate LDV CO targets could 2 vehicles meet stringent CO the transport sector may need avoid 95 Mt in 2030, stronger 2 targets. Efforts to decarbonize to follow an “all-of-the-above” targets could avoid 144 Mt in the grid could further limit the approach for developing cost- 2030 – equivalent to a more indirect emissions from electric- effective low-carbon policies. than 50% increase in emission ity supplied to EVs. benefits. This analysis did not intend to • Biofuels: New biofuel policies are capture all possible policy options • Real-world fuel consumption not an attractive option if only for low-carbon transport, instead gap: On top of the benefits of aiming to reduce direct transport focusing on a more narrow set LDV CO targets, closing the 2 emissions, since biofuels and of technology options for clean gap between laboratory and conventional fuels have similar vehicles and fuels, as well as real-world fuel consumption to direct combustion emission fuel pricing policies due to their 15% could avoid 25 Mt in 2030. factors. Considering fuel lifecycle potential relevance to the EU ETS.

• HDV CO2 standards: Introducing impacts, however, demonstrates Other studies have evaluated the

HDV CO2 standards in 2025 that expanded deployment of 2G potential impacts of additional could avoid 17 Mt in 2030. Early biofuels would have significant measures, including improved infra- introduction in 2020 could avoid benefits compared to 1G biofuels structure for public and non-motor- 55 Mt in 2030 – more than three (26 Mt in 2030), since 2G biofuels ized transport, freight intermodality, times the benefits of a 2025 have lower indirect emissions. and driver training, among others.28 introduction. Some of these measures could be • Combined policies: Implementing • Electrification: Transitioning the most ambitious narrative combined with those evaluated in the LDV fleet to electric-drive considered (narrative ‘6’, which this ICCT study to achieve a 30% vehicles could avoid 19 Mt in includes ‘High LDV efficiency reduction in road transport GHG 2030, even after assuming NTE’, ‘High HDV efficiency’, emissions by 2030. internal combustion and hybrid ‘Biofuels’, ‘Electrification’, and ‘Fuel taxation’) could reduce 28 For example, Ricardo-AEA evaluated the reduction-for-2030_Final_Issue21.pdf direct emissions by 282 Mt potential impacts of a rapid deployment 27 The marginal benefit of fuel taxation alone compared to the baseline in of Communicating Intelligent Transport (‘5’ minus ‘6’) would be somewhat larger if 2030, equivalent to 24% below a Systems (C-ITS) technologies, assuming counted before any other policies, although potential fleetwide efficiency improvements far below the level of reduction targeted for 2005 baseline. Considering the of 0.8% to 1.7% in 2030 depending on the non-ETS sectors in 2030. 30% target for non-ETS sectors, vehicle type (Hill, 2016).

WORKING PAPER 2016-10 INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION 7 REDUCING CO2 EMISSIONS FROM ROAD TRANSPORT IN THE EUROPEAN UNION: AN EVALUATION OF POLICY OPTIONS

Table 4. Detailed CO2 emissions by category and narrative

Cumulative Cumulative Combined percent percent Marginal reduction of reduction reduction

CO2 emissions from LDV reduction in narratives in from 2005 from 2030 Emissions category and HDV [Mt] 2030 [Mt] 2030 [Mt] baseline [%] baseline [%] / Narrative 2005 2030 2030 2030 2030 2030

DIRECT CO2 1. Baseline 891 959 -7.6% 0.0% 2a. ETS only 956 3 3 -7.4% 0.3% 2b. Moderate LDV efficiency 863 93 95 3.1% 9.9% 2c. High LDV efficiency 814 49 144 8.5% 15.0% 2d. High LDV efficiency NTE 790 25 169 11.3% 17.6% 3a. Moderate HDV efficiency 773 17 185 13.2% 19.3% 3b. High HDV efficiency 736 38 223 17.4% 23.3% 4. Electrification 716 19 242 19.5% 25.3% 5. Biofuels 716 1 243 19.7% 25.4% 6. Fuel taxation 677 39 282 24.0% 29.4%

INDIRECT CO2 1. Baseline 352 342 2.7% 0.0% 2a. ETS only 341 1 1 2.9% 0.3% 2b. Moderate LDV efficiency 310 31 32 11.8% 9.4% 2c. High LDV efficiency 294 16 49 16.5% 14.2% 2d. High LDV efficiency NTE 285 8 57 18.8% 16.6% 3a. Moderate HDV efficiency 279 6 63 20.5% 18.4% 3b. High HDV efficiency 266 13 76 24.4% 22.3% 4. Electrification 269 -3 73 23.5% 21.4% 5. Biofuels 243 26 99 30.9% 29.0% 6. Fuel taxation 230 13 112 34.6% 32.9%

DIRECT AND INDIRECT CO2 1. Baseline 1242 1301 -4.7% 0.0% 2a. ETS only 1297 4 4 -4.4% 0.3% 2b. Moderate LDV efficiency 1173 124 128 5.5% 9.8% 2c. High LDV efficiency 1108 65 193 10.8% 14.8% 2d. High LDV efficiency NTE 1075 33 226 13.4% 17.3% 3a. Moderate HDV efficiency 1053 23 248 15.3% 19.1% 3b. High HDV efficiency 1002 51 299 19.4% 23.0% 4. Electrification 986 16 315 20.7% 24.2% 5. Biofuels 959 27 342 22.8% 26.3% 6. Fuel taxation 907 52 394 27.0% 30.3%

Appendix rounded to the nearest million metric well-to-wheel (WTW). Blank cells tons (Mt), some values may not indicate not applicable (for example, A. DETAILED EMISSION exactly match if re-calculated using narrative results are only estimated RESULTS rounded values. Results are reported for years after 2015). The following separately for direct/tank-to-wheel examples may help readers interpret Table 4 includes the detailed emissions (TTW), indirect/well-to-tank (WTT), the results: results that appear in the preceding and direct and indirect combined/ figures. Since emissions estimates are • Baseline direct CO2 emissions are

8 INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION WORKING PAPER 2016-10 REDUCING CO2 EMISSIONS FROM ROAD TRANSPORT IN THE EUROPEAN UNION: AN EVALUATION OF POLICY OPTIONS

Table 5. Direct and indirect emissions by fuel type, including ILUC for biofuels

Percent reduction from conventional fuel Conventional TTW CO2 WTT CO2 WTW CO2

fuel Fuel (g/MJ) (g/MJ) (g/MJ) TTW CO2 (%) WTW CO2 (%) Petrol 73.4 17.2 90.6 Starch Ethanol 71.3 (21.6) 49.7 3% 45% Petrol Sugar Ethanol 71.3 (44.6) 26.7 3% 71% Cellulosic Ethanol 71.3 (73.6) (2.3) 3% 103% Diesel 73.2 18.6 91.8 Vegetable oil-based 76.2 9.2 85.4 -4% 7% Diesel Biodiesel Cellulosic/Waste-oil 72.6 (84.8) (12.2) 1% 113% Diesel substitute Tank-to-wheel (TTW) emissions are commonly referred to as direct emissions. Well-to-tank (WTT) emissions are commonly referred to as indirect emissions. Well-to-wheel (WTW) emissions are equal to the sum of direct and indirect emissions. Values in parentheses are negative.

projected to be 68 Mt higher in Carbon intensity of biofuels the current use of approximately 1.1 2030 than in 2005 (959-891). million tonnes of used cooking oil While direct emissions are the focus in biofuel,32 a conversion efficiency • While electrification (‘4’) would of the ETS, this analysis considers of 0.96 to biodiesel, and a biodiesel have direct emission benefits both direct and indirect emissions in energy content of 37.2 MJ/kg (these of 19 Mt in 2030, these benefits order to account for the fuel lifecycle last two assumptions are from would be offset by 3 Mt due to impacts of low-carbon transport the UK Renewable Fuel Agency’s indirect emissions from upstream policies. Indirect emissions are default value spreadsheet), and an electricity generation. especially important to consider for assumption that the “Cellulosic/ low-carbon fuel policies that promote Waste-oil Diesel substitute” category B. DIRECT AND INDIRECT 1G and 2G biofuels and electrification would account for 25% of all EMISSIONS of road transport. For biofuels, indirect advanced biofuel in the EU. It seems emissions are calculated as the net of reasonable to assume that more Modeling of CO2 emissions production emissions and ILUC minus cellulosic ethanol will be used than Direct and indirect CO emissions comparative fossil fuel emissions. 2 cellulosic renewable diesel because for LDVs and HDVs were Indirect emissions are calculated as: the former is cheaper to produce; estimated using the ICCT’s Global this assumption would result in a Transportation Roadmap model,29 Indirect emissions = (production ratio of 9:2 cellulosic ethanol to which was initially calibrated to emissions + ILUC) – fossil fuel cellulosic renewable diesel. The match IEA’s energy balances for the comparator results of this analysis of direct and EU-28 in 2010 as estimated in 2013.30 indirect emissions by fuel type are For this analysis, historical baseline We took production emission values summarized in Table 5. estimates of combined LDV and HDV from typical emission values from

CO2 were replaced with those from the Renewable Energy Directive In Table 5, the biofuel TTW emission the European Environment Agency Annex V. For starch ethanol we used factors include biogenic carbon

(EEA) after subtracting 11 MtCO2 wheat ethanol (unspecified process emissions; generally, within GHG for motorcycles.31 These revised fuel), for sugar ethanol we used inventory and lifecycle carbon baseline estimates are within 1.5% of sugarbeet, for cellulosic ethanol accounting methodologies, these the previous 2010 baseline. we used waste wood ethanol, for emissions are offset by carbon oil-based biodiesel we used rapeseed sequestration elsewhere in the fuel 29 ICCT (2016, April). Global Transportation biodiesel, and for “Cellulosic/ production lifecycle. TTW values Roadmap Model. Documentation available Waste-oil Diesel substitute” we used from http://www.theicct.org/global- were compared between Argonne transportation-roadmap-model a combination of Fischer-Tropsch 30 OECD/IEA (2013). “2010 Energy Balances.” renewable diesel from waste wood 31 European Environment Agency (2015). and waste vegetable or animal oil 32 Malins, C., Searle, S., Baral, A., Turley, D., “1.A.3.b - Road Transportation.” EEA biodiesel. For this category we and Hopwood, L. (2014). “Wasted: Europe’s greenhouse gas—data viewer. Retrieved Untapped Resource.” Prepared by ICCT and from http://www.eea.europa.eu/data-and- assumed a split of 1/3 biodiesel, 2/3 NNFCC for ECF. Retrieved from http://www. maps/data/data-viewers/greenhouse- renewable diesel. This was based on theicct.org/sites/default/files/publications/ gases-viewer WASTED-final.pdf

WORKING PAPER 2016-10 INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION 9 REDUCING CO2 EMISSIONS FROM ROAD TRANSPORT IN THE EUROPEAN UNION: AN EVALUATION OF POLICY OPTIONS

National Laboratory’s GREET model33 C. FUEL PRICE ELASTICITIES fuel could result in additional environ- and JEC’s Tank-to-Wheels Report.34 mental benefits. The potential impacts of increased The GREET-derived TTW emission fuel taxation were evaluated using factors were taken from the “Results” fuel price elasticities for road D. ADDITIONAL EXPLANATION sheet of the model. The bio-ethanol passenger and freight demand. OF NARRATIVES TTW emissions were assumed to be Passenger elasticities were assumed Narratives are lettered and numbered the same regardless of feedstock. The to apply to LDVs, and freight elastici- to facilitate two kinds of comparison: cellulosic/waste oil mix was a special ties were assumed to apply to vans case, as it was not a fuel mix included (LHDT), MHDT, and HHDT. These elas- 1. Comparisons between lettered in the model. Therefore, a mix of ticities were not applied to buses. scenarios with the same number, 33% soy biodiesel (B100) and 66% Average fuel price elasticity values which represent mutually forest-residue-derived renewable were derived from a literature review exclusive policy choices. For diesel (RD100) was assumed. The of academic studies prepared for the example, a comparison of JEC results were taken from Table UK Department for Transport, but narratives ‘2b’ and ‘2c’ indicates 3-1 of the report, which contained considering results from many other the marginal benefit of setting the main properties of the various countries: these values were -0.3 ambitious LDV efficiency targets fuels considered in the study. The (with a range of mostly -0.1 to -0.5) as opposed to moderate targets. TTW value of bio-ethanol (E100) was for passenger demand and -0.33 inferred via the TTW values of petrol (with a range of -0.25 to -0.4) for 2. Comparisons between scenarios and petrol E10. As with the GREET freight demand.35 While this analysis with different numbers, which values, a mix of 33% biodiesel (FAME) conservatively applied average fuel yield the marginal impact of and 66% Fischer-Tropsch renewable price elasticity values, the authors additive low-carbon transport diesel was assumed. of the literature review identified policies. For example, a elasticity estimates of up to -0.79 comparison of narrative ‘5’ and Grid carbon intensity for passenger demand (pertaining to ‘4’ indicates the marginal benefit To calculate indirect emissions from holiday travel) and -1.07 for freight of new (bio)fuels, assuming

EVs, estimates of grid carbon intensity (from a literature review of papers stringent CO2 standards have using data collected before 1988).36 already been implemented for (grams CO2 per MJ) were derived from IEA’s World Energy Outlook new LDVs and HDVs, and policy (2014). Results were used for the New The percent change in fuel price incentives have increased the Policy scenario to reflect the expec- with a 20 Euro cent per liter tax was uptake of electric-drive LDVs. tation of additional progress (beyond derived from the average fuel price This approach treats most policy the impacts of adopted policies) in in December 2015 according to the options as complementary rather decarbonizing the EU’s power sector. European Environment Agency, and than mutually exclusive, recognizing Converting these estimates from g/ modeled as applying to both petrol that progress needs to be made on kWh to g/MJ resulted in a value of 101 and diesel fuels (rather than each fuel separately).37 many fronts in order to meet GHG gCO /MJ in 2015 that decreases to 66 Given the differential 2 between local air pollutant emissions reduction targets. For example, poli- gCO /MJ in 2030. 2 from petrol and diesel vehicles, cymakers need not choose between

focusing a fuel tax increase on diesel implementing CO2 standards for LDVs or HDVs: both policy options can be

33 Argonne National Laboratory (2015). “The 35 Dunkerley, F., Rohr, C., and Daly, A. (2014). developed provided they are feasible Greenhouse Gases, Regulated Emissions, “Road traffic demand elasticities: A and cost effective. Narratives are and Energy Use in Transportation Model.” rapid evidence assessment.” Prepared numbered roughly according to ease 2015 version 1. Available from https://greet. for Department for Transport by RAND es.anl.gov/index.php Europe. Retrieved from https://www.gov. of implementation, favoring policy 34 Hass, H., Huss, A., and Maas, H. (2013). uk/government/uploads/system/uploads/ options that have a combination of “TANK-TO-WHEELS report version 4.a : attachment_data/file/395119/road-traffic- historical precedent, political appeal, demand-elasticities.pdf WELL-TO-WHEELS ANALYSIS OF FUTURE efficacy and cost-effectiveness; AUTOMOTIVE FUELS AND POWERTRAINS 36 Ibid. IN THE EUROPEAN CONTEXT.” Joint 37 European Environment Agency (2016). however, the order of implementa- Research Centre of the European “Fuel prices.” Accessed 4 Apr 2016 at tion does not particularly matter Commission, EUCAR and CONCAWE. http://www.eea.europa.eu/data-and- when comparing the net impacts of Available from http://publications.jrc. maps/indicators/fuel-prices-and-taxes/ ec.europa.eu/repository/handle/JRC85327 assessment-5 multiple additive policy options.

10 INTERNATIONAL COUNCIL ON CLEAN TRANSPORTATION WORKING PAPER 2016-10