Inducing modal shift from road to rail and waterborne freight is a political objective in many countries [1
], often motivated by ambitious emission reduction, sustainability, and traffic safety targets. In most European countries, modal shift ambitions are particularly driven by the European Commission’s Transport White Paper [2
], while in Norway, modal shift has additionally been a transport-political objective in all National Transport Plans (NTPs) since 2002, state budgets since 2005, and several government agreements [3
Despite this focus, modal shift objectives both in Norway and many European countries are far from being met in practice [4
]. In fact, in most countries, road transport has been increasing (often strongly), while freight has been shifting away from rail to road. Since the 1970s, rail freight’s market share has for example steadily decreased throughout Europe [6
], and particularly in Eastern Europe [5
Important explanations behind these developments include several megatrends strengthening the position of transport by road/truck. A first example is a decades-long trend of sectors starting to organize themselves differently; less nationally and more internationally, country-overspanning, or Pan-European. This trend has caused an increase in international transport, often favoring road transport, as many production and consumption centers can only be reached by road [9
Alongside, debates and developments have been ongoing towards increasing vehicle weight and length allowances in a number of European countries [12
]. While making road transport more cost and environmentally efficient, this development also improves road transport’s competitiveness vis-à-vis other modes and makes shifts to other modes less attractive.
Thirdly, Eastern-European transport firms have taken over sizable shares of international road freight transport in Europe, being able to compete with driver wages below the minimum in many richer EU-countries [5
]. Also this has strengthened road transports’ competitive position.
For the Nordic region (in this article, focusing on Norway, Sweden, and Denmark), similar developments have been visible—transport-generating enterprises such as distribution centers and logistics warehouses are increasingly established at a Nordic level and located in, e.g., the South of Sweden, the share of transport with drivers from low-wage countries has been strongly increasing at the expense of the driver share from the Nordic countries, and vehicle dimension allowances have increased in terms of length, weight, and the use of European Modular Systems [8
In light of the above developments and far too little progress in reducing CO2
emissions from freight transport [8
], a recurrent theme with both policy makers and scholars has been how more modal shift can be achieved than has so far been the case.
For Norway, a freight analysis prepared for the National Transport Plan 2018–2029 [15
] highlighted that assessments of domestic modal shift might underestimate the full modal shift potential, arguing that if more of the imported freight enters Norway by rail or sea; this increases the likeliness of further domestic transports by these modes (rather than by road). As such, the question was posed whether measures implemented at the Nordic level can contribute to increasing the share of foreign freight to and from Norway by sea or rail.
The current study takes a comprehensive approach, assessing a number of scenarios where existing policy measures with modal shift relevance are strengthened or expanded. The policy scenarios studied cover both single, mode-specific policy instruments, as well as combinations of instruments, including cross-border harmonization in Norway, Sweden, and Denmark. As such, we gain important insights into which policy measures might be more effective than others, whether measures might complement each other, and whether international harmonization might increase effectiveness, and thereby modal shift. In addition to effects on transport and modal distribution, we calculate environmental effects through energy use, emissions of CO2 and NOx, and particulate matter. As environmental effects are mode-specific, the latter also has implications for policy making, depending on the trade-offs between for example local and global pollution.
Although this study looks at policy instrument usage in the Nordic region as a whole, it should be noted that analyses are primarily carried out from a Norwegian perspective. Quantitative estimates of changes in modal choices and freight flows are for example made using the National Freight Model for Norway (NFM) [16
], and given its inputs, the analysis only covers effects for freight flows with origin and/or destination in Norway. In addition, future policy scenarios studied with the NFM were specified based on the mapping of policy measures in the Nordics, but also with focus on freight flow analyses with (particularly) Norwegian relevance.
However, introducing modal shift from road transport to sea or rail in Norway’s foreign trade will also lead to less road transport in other Nordic countries, being important transit countries for Norwegian import and export by road. Even though differences in geographical conditions and availability of transport modes will affect how comparable scenarios would turn out in other countries or regions than the Nordics, results and particularly insights from the current study will therefore be highly relevant for researchers, policy makers, and other stakeholders.
The analyses in this study are divided into two stages. First, we use a National Freight Model for Norway to simulate a set of policy measure scenarios outlined in Section 3.3
and their influence on modal choice. Second, these estimates on changes in modal choice are used to compute environmental effects, using data on the fleet for trucks, ships and trains, together with emission factors from updated sources.
3.1. The National Freight Model for Norway
The National Freight Model or NFM for Norway can be classified as strategic transport network model [30
], consisting of the following elements:
Transport demand, represented by commodity flow matrices between Norwegian municipalities, and between Norwegian municipalities and geographical zones abroad, distributed over 39 commodity groups representing different requirements regarding transport quality and time. The commodity flow matrices represent the annual sum of commodity flows between suppliers (producers, importers, and wholesalers) and end-use sectors (exporters, wholesalers, retailers, and service industry) [18
A network model, representing each mode’s physical infrastructure (road, sea, rail, and air) by distance and transport time, including locations of terminals for consolidation and reloading between modes [16
]. There is also one node file for each commodity group, describing the properties for each node and terminal in the model.
Cost functions representing time- and distance-dependent costs for different transport modes, including loading/unloading/reloading, ordering, storing, commodity time values, etc. [17
Optimization routines for choice of shipment size, frequency, and mode, based on a minimization of yearly logistics costs [31
Combining these elements, the model determines optimal transshipment locations (from a list of available terminals) for each pair of origin and destination zones, and then calculates shipment size and transport chains (number of legs, selection of modes and vehicle types). Based on this, the model selects the transport chain with the lowest logistics costs. Model programming is done using an object Pascal Delphi compiler, and due to the involvement of common developers, the NFM features large similarities to the Swedish freight model used by Vierth et al. [32
Policy scenarios assessed in this study are operationalized through changes in parameters in the NFM (e.g., cost parameters, changes in available terminals, costs related to train length restrictions in the network model, etc.). These changes affect which transport chains are calculated to yield the lowest logistics costs, and thus are selected in the model. As such, the NFM yields estimates on the distribution of mode of transport (compared to a reference scenario or “status quo”), impacts on transport costs, and overall logistics costs.
The scenarios are run for the future year 2030, based on projections made for the Norwegian Ministry of Transport and Communications towards the National Transport Plan 2018–2029 [33
]. These projections were based on Statistics Norway’s population projections from 2016 and macroeconomic growth trajectories compiled by the Norwegian Ministry of Finance [34
] with macroeconomic model DEMEC. Growth paths are regionalized using PINGO, a spatial computable general equilibrium model for Norway [35
]. The changes made in the NFM to represent a given scenario compared with a reference situation, are implemented in either the cost functions and the input files generated from those, in the files representing the different nodes, or in the input files for the networks. Beyond the changes and assumptions mentioned in our descriptions, cost developments are assumed to remain the same between modes. Relative costs can change if e.g., the phase-in of new technologies follows a different path or policy changes in levies/duties (e.g., on fuel) are introduced. Such hypothetical situations were not assessed, in order to be able to distinguish the partial effects of modal shift measures.
3.2. Environmental Effects
NFM-output on transport performance (tonne-km) from domestic and foreign trade was combined with energy use and emission factors to estimate environmental effects in the different policy scenarios (changes in energy use, GHG- and NOx emissions, and exhaust particulate mass (PM)). Emissions and energy use factors (in g/tonne-km and MJ/tonne-km respectively) were derived from different sources, depending on transport mode.
For road transport, energy use and emissions were calculated using transport performance output, together with fuel consumption and emission factors from the HBEFA-model (v.3). Both the distribution of road transport over different-size trucks and maximum load capacity for transport of different commodity groups are output from the NFM. The NFM divides trucks into light lorries, heavy lorries, and large trucks, each with several sub-categories. These subcategories were matched with truck sizes in HBEFA. Based on a weighting of sub-categories (an output of the NFM), energy use for each main truck category was calculated. Average load factors were estimated using information from the NFM on load capacities for each commodity group, to take into account that different commodities have different volumes per tonne. When applying environmental emission factors, empty trips were assumed to constitute 30% of the total distance, derived as average for freight transport with trucks in statistics for 2016–2018 from Statistics Norway.
For emission assessments of Norwegian freight trains, we used the basic methodology from energy and emission calculation system EcoTransIT [36
]. In EcoTransIT, calculations are based on train weight (gross and net), after which conversion factors are used to correct for the positioning of empty wagons and topography. This work describes rail transport with three main train types, respectively “wagonload,” “other rail,” “diesel trains,” and their sub-categories from the NFM. To calculate typical train weights within each sub-category, maximum capacities (from the NFM) were combined with tare weights for freight wagons from EcoTransIT. Average energy use for each main train type was calculated by weighting sub-categories (from the NFM), after which emission factors were derived based on the Norwegian electricity mix employed in EcoTransIT [36
]. This mix was validated against the most recent statistics available.
For sea transport, the NFM uses a set of ship types. Characteristics of these were developed using data for existing ships of the same type and similar size from SeaWeb [37
]. Fuel consumption for each ship type was obtained using specific fuel consumption data from IMO [38
] together with average speeds from SeaWeb. Emission factors for calculating air pollutants emissions were based on Cooper and Gustafsson [39
The share of renewable diesel blend in 2030 is assumed to be 20%, both for road transport and diesel trains. This is a continuation of the mandatory blend in 2020, as shares in later years have not been decided yet. Further, we assumed that all trucks used in 2030 comply with Euro-VI emission standards, while the share of fully electric trucks or electric hybrids is expected to be negligible [following 40]. For ships, fuel is assumed to remain of 100% fossil origin in 2030, and energy-efficiency and emission standards are held at current levels. Finally, all emissions are presented as tailpipe emissions and do not include upstream emissions from e.g., production and transportation of fuel, and elements of uncertainty discussed in the discussion section.
Based on a mapping of existing policy measures for modal shift in the Nordic countries (an update of Pinchasik et al. [8
], analyses of volume flows and developments in foreign trade with trucks [14
], and feedback on modal shift potential from a survey among firms with own sea terminal [40
], we constructed a set of future policy scenarios. The first four scenarios consider single policy measures, while the latter five scenarios consider combinations of the single measures and/or border-crossing implementation in the Nordic countries. Together, the scenarios cover both infrastructure/efficiency improvements and financial incentives (in line with [5
]) and can be characterized as both push and pull measures [26
]. Vierth et al. [32
] carried out a related analysis as in the current paper for the case of Sweden, using the Swedish national freight model to calculate modal shifts, environmental effects, and changes in logistics costs from several proposed policy measures. Policy measures were studied in isolation and combined to illustrate whether measures reinforce each other or require coordination.
Both the scenarios covering single policy measures and those covering combinations and border-crossing implementation are assessed against a reference scenario based on the latest NFM, with 2030 as year of comparison (i.e., with road network and program versions of March 2019 and PWC-matrices for 2030 freight flows). The reference scenario represents “business-as-usual” and does therefore not include eco-bonuses or changes to current road pricing regimes. For railway, it includes current length allowances of 480 m for combi-trains (containers) and wagon loads, except for some border-near rail stretches where length allowances vary around ca. 600 m.
3.4. Scenarios Covering Single Policy Measures
3.4.1. Scenario 1: Norwegian Ecobonus for Sea Transport
In 2017, Norway established a trial scheme for an “ecobonus” for sea transport. Subject to certain requirements, this scheme was designed to pay out support towards sea transport replacing road transport on Norwegian territory. In May 2019, Norwegian Government [24
] proposed to replace the scheme by other measures, but later announced to make the scheme permanent with an annual budget of NOK 50 million [42
Due to the frequent recent changes, the current policy scenario is assessed for an ecobonus scheme with hypothetical budget of NOK 150 million/year, equal to the rail ecobonus in Scenario 2.
To implement the sea ecobonus in the NFM, we assume that the entire annual budget is paid out to shippers. Based on experience from practice, we further set eligibility limitations, including all ships suitable for transport of general cargo, but excluding, e.g., bulk transport and comparable (which already goes by ship where possible).
Based on the 3.9 million tonnes of goods that, from the latest NFM version, are loaded/unloaded from container ships in Norway annually, and considering that freight is handled twice in terminals, the ecobonus was operationalized as reduction in terminal costs for loading/unloading in Norwegian ports. Because the NFM does not allow applying cost reductions exclusively to newly generated sea transport, and some modal shift could be expected from road to sea due to reduced terminal costs, exact terminal cost reductions were calculated in an iterative process. Cost reductions per tonne were set equal for eligible ship types. Further, if the calculated terminal cost reduction yielded total support payments above/below the program’s budget, parameters were adjusted, and the model reiterated. Assumptions for this scenario were implemented as direct changes in the port costs in the nodes files.
3.4.2. Scenario 2: Norwegian Ecobonus for Rail Transport
Until recently, when Norway’s revised State Budget for 2019 included a “rail ecobonus” proposal, Norway did not have an ecobonus program for rail transport, such as, e.g., Sweden, or previously Italy [43
]. We therefore modeled a policy scenario with an annual “rail ecobonus” budget of NOK 150 million, i.e., equal to the sea ecobonus in Scenario 1. Again, support is assumed to be payable only for the parts of road transport on Norwegian territory that are shifted away. Support is further assumed to be limited to combi-transport (i.e., excluding bulk and timber transport).
The ecobonus for rail was implemented in the NFM in a similar way as for the sea scheme, here with a reduction in terminal costs of 15 NOK/tonne combi-freight handled in Norwegian rail terminals (at both ends of the transport chain). The assumptions for this scenario were implemented as changes in the terminal costs in the nodes files.
3.4.3. Scenario 3: Increases in Eurovignette Rates in Sweden and Denmark
With regard to costs for trucks driving in the Nordic countries, we started out using the costs of today’s Eurovignette, which is used in Sweden and Denmark (in addition to Luxemburg, the Netherlands, and until recently, Belgium), and applies to heavy trucks (gross weight ≥12 tonnes). For this policy scenario, we assumed a five-doubling of today’s (daily) Eurovignette price, i.e., 5 × 8 EUR. To reflect that the Eurovignette applies in Sweden and Denmark, but not Norway, this rate increase is operationalized through the modelling of a “toll charge” for driving into or out of Sweden or Denmark. In the model, the assumptions for toll charges were implemented in the network files.
Although the rate increase assessed here might seem large, Bouchery and Fransoo [23
] find that while implementing taxes on road transport aligns costs and CO2
emissions, cost increases for the road mode have to be substantial, often beyond what is considered feasible in practice, to have significant effect. For example, for truck transport between Oslo-Bergen or Oslo-Trondheim in Norway, toll expenses (one way) average around 360 NOK/36 EUR (for Euro VI trucks, outside of peak hours).
3.4.4. Scenario 4: Longer Freight Trains
In order to lower operational costs and freight rates for rail, particularly important factors are the ability to operate heavier, longer and wider trains, with higher speeds and better capacity utilization [19
]. Such factors can improve rail’s competitiveness versus road by reducing unit costs per tonne-km. However, the extent of cost reductions depends on whether longer trains require changes to locomotive set-ups, such as extra locomotives or by replacing four-axle locomotives (used by most operators on the Norwegian freight network) by more powerful six-axle versions.
For this scenario, we analyze effects of facilitating longer freight trains for transport to/from Norway at Kornsjø and Charlottenberg (see Figure 1
). Train length allowances are set to 740 metres for combi-trains (the required allowed length on the Trans-European transport network, TEN-T, increasing from the current ca. 600 m). We further assume length allowances of 640 m on main freight relations in the Norwegian rail network, an increase from 480 m for combi-trains (containers) and wagon loads. Where most cost-effective, we assume operation of six-axled locomotives. Finally, we take into account that for longer trains to yield modal shift, freight flows in the market must be large enough to fill up the extra train capacity at the same level of service (frequency). All these changes were implemented as direct input into the NFM’s cost model and thereby the generated cost input files.
3.5. Combinations of Measures and Border-Crossing Implementation
For the following scenarios, we ran combinations of different policy measures in conjunction, and with implementations not just in Norway, Sweden, or Denmark, but in multiple countries at once. This is done to assess whether measures might complement each other and whether international harmonization might increase their effectiveness and resulting modal shifts.
3.5.1. Scenario 5: Combination of Longer Freight Trains and Norwegian Ecobonus for Rail
This fifth policy scenario combines increased freight train lengths (with operationalization as in scenario 4) with a rail ecobonus in Norwegian terminals only (as in Scenario 2).
3.5.2. Scenario 6: Idem Scenario 5, but with Rail Ecobonus also Applying in Swedish and Danish Rail Terminals
This scenario follows the fifth scenario above but assumes that the same rail terminal cost reductions as in Scenarios 2 and 5 also apply to Swedish and Danish rail terminals in the model’s network (implicitly assuming that these countries make available sufficient public funds to accommodate this). The reasoning behind this is that Norwegian and Swedish schemes respectively, provide ecobonus support based on parts of transport carried out on each country’s own territory. If transport firms or customers could receive support for the entire distance that goods are transported by rail instead of by road, this would cover a larger part of total shipping costs and (theoretically) increase the probability for modal shift.
3.5.3. Scenario 7: Combination of Road Measures, with Rail and Sea Measures in Norway
In this scenario, several policy measures are combined. We simultaneously consider the ecobonuses for sea and rail transport from Scenarios 1 and 2 (thus applying in Norway only), and the Eurovignette rate increases in Sweden and Denmark, from Scenario 3.
3.5.4. Scenario 7b: Expansion of Scenario 7 with Terminal Cost Reductions in Sweden and Denmark
To better understand effects on modal choice, we ran an additional simulation of Scenario 7, but expanding cost reductions for sea and rail as in Scenarios 1 and 2 from Norwegian, to also apply in Swedish and Danish ports and terminals. For this scenario, environmental effects were not calculated.
3.5.5. Scenario 8: Combination of Road, Rail, and Sea Measures, Coordinated for the Nordics as a Whole
In this final scenario, several policy measures are combined and coordinated for the Nordics as a whole. We consider ecobonus-induced cost reductions in both Norwegian, Swedish and Danish ports and terminals in the model’s network, together with cost increases for road freight. For the latter, we assume that increases in Eurovignette rates as in Scenario 3 (or similar cost increases) also apply within Norway. This is operationalized as increase in per-km costs for semitrailers and European Modular Systems (25.25 m vehicles) of 0.60 NOK/km in all three countries. Given an annual mileage of 100,000 km this is equal to a five-doubling of today’s Eurovignette costs. The increase is also equal to rates in a Swedish road pricing proposal as Eurovignette replacement [44
], as was recently rejected by Swedish Parliament and has also been analyzed in Vierth et al. [32
3.6. Scenario Overview
To summarize, Table 1
provides an overview of the different policy scenarios and their main assumptions.
5. Discussions and Conclusions
5.1. Induced Modal Shift Through Combinations of Policy Measures
In light of underachievement on modal shift objectives and too little progress in reducing CO2
emissions from freight transport given climate commitments [8
], a recurrent theme with both policy-makers and scholars has been how more modal shift can be achieved than has so far been the case. In this article, we analyzed the effects of strengthening existing policy instruments for transferring freight transport from road to sea and rail, and of harmonizing policy-instruments across borders between Norway, Sweden and Denmark, for potentially more effect.
From our simulation of policy scenarios, we find that a Norwegian ecobonus scheme for rail yields much larger modal shifts away from road transport than a similar ecobonus for sea transport. The rail ecobonus also yields positive environmental effects, with small reductions in emissions of CO2,eq, NOx and PM, rather than minor increases under the sea ecobonus scheme. This is due to the high degree of electrification of the rail mode, and because with the sea ecobonus, more transport performance is added on sea than reduced on road, due to longer distances stemming from the location of the transferred goods. Further, most ships have higher specific NOx and PM emissions relative to modern trucks.
Significant increases in Eurovignette rates in Sweden and Denmark also result in reduced road transport, mostly through shifts from road to rail and particularly by affecting road route choices between southern and northern Norway, shifting from transit through Sweden to transport within Norway. As a whole, however, modal choice and environmental effects are limited. Allowing longer freight trains, in turn, has a larger impact on modal choice than the ecobonus schemes and road cost increases, and also yields larger decreases of environmental emissions. However, the infrastructure investments required are expected to entail much higher policy costs for government [29
Combining a Norwegian rail ecobonus with policy measures facilitating longer trains yields slightly stronger effects than the sum of effects when the measures are studied in isolation, by causing some additional shifts from sea to rail. However, further expansions, by applying ecobonus-induced cost reductions also in Swedish and Danish rail terminals, only result in a marginally larger reduction in road transport. This suggests that harmonizing the rail ecobonus across the Nordic countries yields little additional effects, despite requiring the use of more public funds. Most likely, this is due to distances between the Nordic countries being too short to exploit rail transport’s main benefits and to overcome inherent weaknesses, such as (expensive) road dependency at origin and/or destination and long lead times [6
Combining both sea and rail ecobonuses in Norwegian ports and terminals with Eurovignette rate increases in Sweden and Denmark gives strong modal shifts from road to rail, and smaller shifts to sea transport. Because of different distances for different modes on different relations, tonne-km increases for rail and sea transport in sum are larger than the reduction in tonne-km by road, as a result of the relations with changes in mode choice. While the increased use of rail results in reduced CO2,eq emissions, emissions of air pollutants increase due to increased use of ships. In many cases, however, sea transport will have shorter distances than land-based transport between Norway and the European Continent, suggesting that effects of measures might be case-specific, and that emissions in many situations could decrease.
When additionally expanding sea and rail ecobonuses to apply also in Swedish and Danish terminals, results suggest that such harmonization yields additional modal shift away from road, particularly to rail. Simultaneously, such expansions of both rail and sea ecobonuses seem to increase rail’s competitiveness versus sea transport, resulting in transfers from sea to rail. This contrasts findings from Scenarios 5 and 6, where harmonization of the rail ecobonus yielded only marginal additional effects.
Finally, a harmonization of ecobonuses for both sea and rail, combined with a per km-charge for road transport in all three Nordic countries, results in minor additional modal shifts compared to the previous scenarios. This is followed by a decrease in CO2,eq emissions and somewhat larger decreases in emissions of air pollutants.
Our results suggest that in some, but not all cases, harmonization of policy measures such as ecobonuses may strengthen effects of modal shift policy, depending on transit traffic and border-crossing effects.
5.2. Environmental and Other Considerations
Altogether, most scenarios show reduced environmental emissions, particularly when policy measures are combined and/or include the facilitation of longer freight trains. However, even in scenarios with rather strong policy measures, reductions in CO2,eq
emissions do not exceed 3.6% in 2030. This indicates that modal shift can only be a moderate contributor to the decarbonization of freight transport, and is in line with observations by e.g., Tao et al. [1
], Kaack et al. [5
], Pinchasik et al. [8
], and McKinnon [11
Moreover, in several scenarios, we find increased air pollution. This is due to increases in sea transport, which has higher specific emissions of NOx and PM. Policy makers aiming at modal shift should therefore also consider other environmental impacts than CO2
. Further, in line with Bouchery and Fransoo [23
], maximizing modal shift is not necessarily optimal for reducing CO2
emissions. As intermodal transport in some cases increases transport distances compared to direct truck transport, environmental effects of modal shift depend on a trade-off between efficiency gains and losses due to longer transport distances (as well as mode-specific changes).
However, it should be noted that even when modal shift measures do not yield large effects for society as a whole, governments may find them desirable for other reasons. An example from Norway includes a fear that without measures, even more rail freight routes might be cancelled. Moreover, modal shift may be considerable in specific transport corridors where rail or sea is desirable. Further, even when emissions and mode choice for the whole of transport chains do not change much, countries might have an interest in what happens on their own domestic territory.
5.3. Policy Design, Developments, Assumptions, and Uncertainty
Lessons learned from the analysis are that support schemes such as the ecobonuses for rail and sea must be designed such, that they are only paid out if the support results in a (new) modal shift away from road transport. Based on Norwegian freight flows, we find that harmonizing over the borders of the Nordic countries would do this in some, but not all cases, when it comes to Norwegian freight flows. However, what is not analyzed here is the impact of the assessed policy measures on Swedish and Danish freight flows, which can make it possible to establish new shipping or rail routes that can attract goods from the Nordics as a whole. In this regard, implemented policy measures should be evaluated and findings shared between the Nordic countries, as this may improve the design of new policies.
Given this study’s conclusions, it is appropriate to remember that model simulations do not capture all societal trends. There are at least three megatrends that strengthen the competitive position of road transport and trucks [14
]: Establishment and use of Nordic distribution centers, increased use of transport firms from lower-wage countries for border-crossing transports, and increases in vehicle dimensions in terms of weight and length [8
]. These developments have improved the competitive position of road transport and are particularly relevant because wage expenses constitute a larger share of total transport costs for trucks than for rail or sea transport. Policy makers should therefore take into account that, while measures such as larger allowances for vehicle dimensions might improve the efficiency of road transport, they also make road transport more competitive versus rail and sea.
As all projections about the future, developments in important drivers and assumptions are subject to uncertainty. In our analysis, important factors are particularly the employed emission factors, commodity flows and projections of these toward 2030, and an assumption of equal developments in future costs for the different transport modes.
With regard to emission factors, we for example assumed that electric trains use a Norwegian electricity mix, which, due to its renewables share, implies relatively large environmental benefits (both CO2
and air pollutants) when shifting from sea/road to rail transport. Different electricity mixes, e.g., based on electricity production using natural gas, would yield smaller, albeit still positive environmental benefits from modal shifts to rail. Similarly, we assumed that the share of electric trucks is still small by 2030 [40
]. A large-scale introduction of electric or significantly more efficient vehicles (e.g., resulting from recently adopted emission standards for heavy-duty vehicles [47
]), or a higher blend-in requirement of biofuels, would make the environmental benefits of modal shifts away from road, smaller.
Regarding cost development, deviations from our assumptions can affect the competitive position of different modes, and where this competitiveness change is large enough, result in a mode change. Examples could include cost increases (decreases) from a different technology uptake track, further increases in vehicle dimensions, regulation disproportionally affecting one mode (e.g., emission requirements), fuel requirements, levy increases on fossil fuels, etc.
Finally, commodity flows may develop differently than projected, both in volume terms as well as with regard to origins/destinations and relative changes between commodities. This may have implications for how much transport takes place, where, and by which mode. Depending on the deviation from projections, this may imply both increases and decreases in transport performance overall, and for different modes. The same applies to environmental effects. Altogether, uncertainty may therefore have an impact on results. For many factors, this impact will likely not be very large.