Aarhus University Seal

No. 252: Emissions from shipping in the Arctic from 2012-2016 and emission projections for 2020, 2030 and 2050

Winther, M., Christensen, J.H., Angelidis, I. & Ravn, E.S. 2017. Emissions from shipping in the Arctic from 2012-2016 and emission projections for 2020, 2030 and 2050. Aarhus University, DCE – Danish Centre for Environment and Energy, 125 pp. Scientific Report from DCE – Danish Centre for Environment and Energy No. 252 http://dce2.au.dk/pub/SR252.pdf

Summary 

Navigation is one of the most important local sources of emissions in the Arctic area. Navigation emissions are characterized by the fact that the air pollutants formed during ship engine combustion are being injected directly into the Arctic environment along the vessels route at low vessel chimney heights. Hence, the emission deposition from vessels sailing in the arctic area becomes relatively large compared to the total emissions emitted by these vessels (e.g. Corbett et al., 2010; Winther et al., 2014). In this respect, the emissions of black carbon (BC) are of particular interest. It is well known that BC has global warming properties due to its ability to absorb light over reflective surfaces e.g. snow covered surfaces and due to its darkening effect when deposited to snow and ice surfaces (e.g. Quinn et al., 2008; Flanner, 2007, 2009). 

In a previous ship emission project, carried out by DCE – Danish Centre for Environment and Energy at the Institute for Science and Technology at Aarhus University, a spatial distributed emission inventory for the Arctic area above 58.95N was made for the year 2012 based on satellite AIS data, ship engine power functions and technology stratified emission factors (Winther et al., 2014). Emission projections, emission concentration and deposition calculations were also made for the years 2020, 2030 and 2050, and further calculations were also made by combining the baseline projected emissions with additional ship emissions estimated for polar diversion routes due to the possible extent of polar sea ice in the future. 

This report presents the results from the DANCEA project “Emissions from shipping in the Arctic from 2012-2016 and updated emission projections until 2050” carried out by DCE – Danish Centre for Environment and Energy at Department of Environmental Science at Aarhus University with financial support from the Danish Cooperation for Environment in the Arctic (DANCEA). The historical emission estimates now cover the historical years 2012-2016 for the Arctic area above 58.95N, and the emission projections for the years 2020, 2030 and 2050 are improved by using the more robust five year weighted historical data for ship traffic, updated business as usual (BAU) and high traffic growth (HiG) rates and emission factors. Emission factor adjustments for ship engine load are introduced in the calculations as well as updated assumptions for fuel type usage and the deployment of EGCS (Exhaust Gas Cleaning System) for SOx scrubbing in the projection years. For historical and projected emissions, the report also presents results for the concentration and the deposition of the emissions. Updated ship emission estimations along polar diversion routes are also presented in this report. 

In addition to the Baseline emission projections, this report also presents emission results for a SOx Emission Control Area (SECA) scenario and a heavy fuel oil (HFO) ban scenario. In the SECA scenario, the existing SECA zones (i.e. America and North Sea/Baltic Sea SECA’s) are expanded to cover the entire inventory area. In the HFO ban scenario, no use of HFO by ships is allowed in the inventory area at all. 

The full list of emission components estimated in the project are the short lived climate pollutants and their precursors SO2, NOx, CO, NMVOC, PM, BC and OC and the greenhouse gases CO2, CH4 and N2O. The main report focuses on the inventory results for fuel consumption, SO2, NOx and BC for BAU traffic growth. Results for the remaining emission components and high growth results are shown in the report annex. 

For 2012[2013, 2014, 2015, 2016], the following total results are calculated for fuel consumption: 4.8[5.1, 6.3, 6.6, 5.4] MTonnes; SO2: 82[84, 108, 60, 53] kTonnes; NOx: 320[339, 429, 432, 361] kTonnes and BC: 0.71[0.73, 0.86, 0.65, 0.56] kTonnes. 

Per ship type in the latest historical year 2016, the largest shares (percentage values in brackets) of fuel consumption (by mass), BC, NOx and SO2 are calculated for fishing ships (45 %, 33 %, 39 %, 7 %) followed by passenger ships (21 %, 22 %, 18 %, 25 %), general cargo (10 %, 16 %, 13 %, 22 %), tankers (10 %, 12 %, 14 %, 21 %), bulk carriers (4 %, 6 %, 5 %, 12 %) and container ships (3 %, 4 %, 5 %, 7 %).   

For the years 2012-2016, the total NOx emissions levels are quite similar to the fuel consumption levels (Figure ES 1). In 2015 and 2016, the calculated SO2 emissions drop significantly due to the stricter fuel sulphur regulations in SECA’s from 1.1.2015 causing a fuel switch from 1.0 % fuel oil to marine diesel oil (MDO)/marine gas oil (MGO) in SECA’s by ships using HFO by origin. The latter fuel switch also brings significant BC emission reductions due to the lower BC emission factors for MDO/MGO compared to 1.0 % fuel oil. 

The SO2 emission reductions in the SECAs from 2012 to 2016 are clearly visible from Figure ES 2. Although less visible, BC emission reductions in the SECAs also appear from Figure ES 2 whereas for NOx there are no clear visual differences in the overall emissions distributions shown for 2012 and 2016. 

In the Baseline scenario for the forecast years 2020[2030, 2050], the following total results are calculated for fuel consumption: 5.7[5.8, 6.3] MTonnes; SO2: 17[17, 18] kTonnes; NOx: 371[318, 247] kTonnes and BC: 0.43[0.44, 0.47] kTonnes. 

The most convenient way to assess the emission results for the Baseline BAU scenario is to explain the emission development from today’s emissions calculated by using the consolidated (five-year weighted average) ship activity data and the uniform grid system used to project the ship activity data and emission factors for 2016 (the so-called uniform gridded 2016 inventory). 

The following percentage changes between the uniform gridded 2016 results and the Baseline BAU results for the forecast years 2020, 2030, 2050, respectively (results in brackets) are calculated for fuel consumption ( +3 %, +5 %, +14 %), BC ( -19 %, -18 %, -12 %), SO2 (-70 %, -70 %, -69 %) and NOx (+1 %, -14 %, -33 %). 

The NOx emission reductions during the forecast period (Figures ES 3 and ES 4) is due to the decrease in NOx emission factors. The spatial NOx emission reductions are most significant for the North SEA/Baltic Sea emission control area (Figure ES 4) where new engines installed on board ships from 1 January 2021 must comply with the most stringent IMO (International Maritime Organization) Tier III NOx emission standards. 

For SO2 and BC, the major reason for the emission reductions from 2016 (uniform gridded) to 2020 shown on Figure ES 3 is the shift from HFO fuel with a sulphur content of 2.45 % in 2016 to 0.5 % fuel oil in 2020 and the consequently reduced emission factors. 

Despite the increase in total fuel consumption from 2020 onwards, the total emissions of SO2 are almost stable during the forecast period due to the increasing amount of HFO being used in combination with EGCS (Figures ES 3 and ES 4). Also the gradually increased liquefied natural gas (LNG) fuel consumption assumed in the Baseline scenario plays an important role in the emission explanation for SO2. The emissions of BC increase somewhat due to higher BC emission factors for HFO in combination with EGCS compared with the emission factors for the fuel being replaced (MDO/MGO inside SECA’s, 0.5 % fuel oil outside SECA’s).  

In the SECA scenario for the forecast years 2020[2030, 2050], the following total results are calculated for fuel consumption: 5.7[5.8, 6.3] MTonnes; SO2: 9.0[9.1, 9.6] kTonnes; NOx: 371[318, 247] kTonnes and BC: 0.42[0.42, 0.46] kTonnes. For the HFO ban scenario the 2020[2030, 2050], results for fuel consumption become: 5.7[5.8, 6.3] MTonnes; SO2: 8.9[8.9, 9.3] kTonnes; NOx: 371[318, 247] kTonnes and BC: 0.40[0.40, 0.41] kTonnes. 

The fuel consumption and NOx emission totals are the same for the Baseline, SECA and HFO ban scenarios due to equal specific fuel consumption and NOx emission factors for HFO and MDO/MGO and similar LNG shares of total fuel consumption per forecast year assumed in all three cases. In all scenario years, the calculated SO2 emissions for the SECA and HFO ban scenarios are almost half of the emissions calculated for the Baseline scenario (Figures ES 5 and ES 6) for the following reasons: in the SECA scenario HFO is only used by ships in combination with EGCS and the 0.5 % fuel oil used outside the original SECA area, and not being replaced by LNG, is being replaced by MDO/MGO. In the HFO ban scenario all HFO consumption by ships not being replaced by LNG is replaced by MGO/MDO. 

For BC in 2020[2030, 2050] the HFO ban and SECA emissions are 8 %[9 %, 12 %] and 3 %[3 %, 3 %] smaller, respectively, than the Baseline results. Apart from LNG with similar fuel consumption shares assumed in all scenarios, in the HFO ban scenario only MDO/MGO fuel is used with a correspondingly low BC emission factor. In the SECA scenario, HFO fuel in combination with EGCS is used with a relatively higher BC emission factor. However, in the SECA scenario, the BC emissions from MDO/MGO replacing 0.5 % fuel oil are smaller than the emissions from 0.5 % fuel oil used in the Baseline scenario due to the level of the BC emission factors. 

In the previous ship emission inventory for the Arctic area made by Winther et al. (2014), results were calculated for the historical year 2012 and for the projection years 2020, 2030 and 2050.  

The inventory changes for 2012[2020, 2030, 2050] becomes -3 %[8 %, 7 %, 5 %] for sailed distance, -5 %[23 %, 22 %, 20 %] for total fuel consumption, -7 %[-44 %, -44 %, -46 %] for SO2, 4 %[23 %, 24 %, 18 %] for NOx and -55 %[-73 %, -74 %, -75 %] for BC. 

There are multiple reasons for the above listed changes in the results. The most important qualitative explanations are given in the following. 

In the previous inventory, the traffic growth rates were based on 2012 ship activity data. In the present inventory, the traffic growth rates are based on consolidated five-year weighted traffic data. This consequently changes the ship activity distribution by ship type and ship size and impacts the total fuel consumption and NOx emissions to the most extent. 

For SO2, the emission changes between the previous and the current inventory are predominantly due to adjustments made in fuel sulphur content for MDO/MGO as input for the emission calculations in the present inventory. The introduction of LNG fuel consumption in the current inventory also induces differences in the inventory results and the increasing amount of LNG fuel used in the projection years implicitly brings down the total emission estimates of SO2 and BC in the present inventory. 

For BC, the most important emission changes originate from the changes in emission factors. In this way the previous BC emission factor of 0.35 g/kg fuel regardless of fuel type has been changed in the current inventory to a set of much lower emission factors for MDO/MGO, 2.5 % HFO and 0.5 % fuel oil. Despite being adjusted upwards due to relatively low ship engine loads, the aggregated BC emission factor in the present inventory becomes 0.15 g/kg fuel for the entire area in 2012. 

The emission factors used in this work are in agreement with recent BC emission measurement conclusions from the international project on maritime BC emissions coordinated by the International Council of Clean Transport (ICCT, 2016), and important BC emissions review work made by Lack and Corbett (2012) and Lack (2016). However, due to the well-known scarcity of BC emission data for ship engines as input for emission inventories, the calculated results for BC emissions must be regarded with care. Any new data arriving on BC emission factors will be carefully assessed in future shipping emission inventory projects. 

The average surface concentration and BC deposition increases from navigation (Figure ES 6) remain on low levels for the baseline scenarios without diversion traffic included. For the projection scenarios with additional diversion traffic, the forecasted increases in surface concentration and BC deposition become far more pronounced. In both cases, with or without diversion traffic, the surface concentration/deposition increases become significantly higher during summertime. Solely for diversion traffic, the summertime average surface concentration/deposition percentage increases for BCcon[BCdep, SO2, O3] become 0.8 % [0.2 %, 1.1 %, 3.3 %] and 3.9 % [1.1 %, 8.3 %, 8.7 %], respectively, for BAU and HiG in 2050. And solely for diversion traffic, the yearly average surface concentration/deposition percentage increases for BCcon[BCdep, SO2, O3] become 0.2 % [0.1 %, 0.1 %, 0.8 %] and 1.0 % [0.4 %, 1.0 %, 2.1 %], respectively, for BAU and HiG in 2050. 

Also the spatially distributed average navigation contributions are low in most of the inventory area in 2016 (Figure ES 7). The additional contribution to BC surface concentrations, however, reach up to 5 % mainly around Iceland, and high SO2 additional contributions (20-100 %) are calculated in some sea areas due to reasons explained earlier. In 2050, during summertime, the additional contributions from navigation (mainly from diverted traffic) becomes very visible for BC (20-40 %) and SO2 (300-1000 %) along the Arctic diversion routes, while the BC deposition (1-2 %) and the O3 concentration (7-10 %) and become highest over the oceans east of Greenland and in High Arctic, respectively.