Area: Economic Modeling and Policy
Problem class: MCP

Modeling Transportation Carbon Intensity Targets for the EU with GAMS


The International Council on Clean Transportation (ICCT) is a non-profit organization, helping governments and policy makers make the right decisions to reduce air pollution and reduce fuel consumption, across all modes of transport. Given the complexity of the transportation sector, policy makers can only make informed decisions, if the policy options can be simulated using a consistent set of rules using complex, integrated models of the whole sector. Algebraic modeling languages (AMLs) such as GAMS have proven to be useful tools for developing such models. For a project on decarbonization of the transport sector, initiated by the European Commission, the ICCT asked GAMS to develop, test, and run a partial equilibrium model of the transportation sector in the EU, including light duty vehicles, heavy duty vehicles and a representation of the aviation sector. The purpose of the model was to test several policy scenarios that included overlapping green-house gas (GHG) reduction targets, eligibility caps, and other preferential treatments. Within GAMS we can apply all of these constraint sets simultaneously and solve the model to gauge the market’s response to these policies. The results of this work feeds into the EU commission's 'Green Deal' plan on climate change, with the aim of reducing greenhouse gas emissions by 55% by 2030, and becoming climate neutral by 2050. The following text summarizes some aspects of the study in a condensed format. The full report can be accessed on the ICCT website .

Considered Policy Scenarios

A total of 10 policy scenarios are considered, which were developed in collaboration with the researchers at the ICCT. Each scenario represents a combination of:

  • Greenhouse gas reduction targets or renewable energy mandates
  • Caps on food-and-feed based biofuels
  • Advanced biofuel mandates
  • Renewable fuels of non-biological origin (RFNBOs)
  • Sustainable aviation fuel (SAF) mandates
  • Aviation e-fuel mandates
  • Expected electric vehicle annual growth rate
  • Caps on use of certain intermediate crops (e.g. soy or maize) for biofuel production

The scenario parameters are summarized below:

Study Scenarios
Table 1. Input scenarios

The Model

The model was formulated as a static partial equilibrium model, using multiple agents. Each of these agents acts as a cost minimizer, while complying with the set of proposed policy targets:

  • Consumer agents make purchasing decisions for Light Duty Vehicles or Heavy duty vehicles with different engine technologies (gasoline, diesel, electric, hydrogen fuel, or compressed natural gas).

  • An aviation consumer agent makes purchasing decisions about the fuel blend to purchase.

  • A blender agent, responsible for providing fuel to the consumer agents, meeting the policy requirements. The supply of blendstocks to the blender is assumed to fit an iso-elastic supply curve.

About Extended Mathematical Programming (EMP)

Equilibrium problems can be tedious to implement, but GAMS – as the only algebraic modeling language – offers the “Extended Mathematical Programming” (EMP) extension with special support for these problems. EMP makes it possible to implement the individual agents as separate subproblems, which are then automatically reformulated into a format that can be solved efficiently by commercial solvers. The true power of this reformulation technology is that it enables rapid model re-development without concern for necessary staff time associated with the reformulation-debug-verify cycle. This cycle of re-development could quickly become overwhelming if late changes were necessary, something that is inherent in a project that aims to inform a constantly evolving policy conversation.


Figure 1 summarizes the types of fuels used to meet the policy requirements for each scenario. Some of the differences between the scenarios are as expected, for example a lower total amount of renewable energy when the GHG target is reduced (Scenario 2) and no food-based biofuels when the food-based cap is set to 0% (Scenarios 3, 4, and 10). One striking result is the large amount of intermediate crop biofuel in most scenarios in which it is exempt from the food-based biofuel cap. When the policy becomes more ambitious, for example increasing the energy mandate level from Scenario 8 to 9, intermediate crop biofuel fills in most of the total increased renewable fuel demand. In particular, we find a large increase in soy hydrotreated vegetable oil (HVO). Simply reducing the target level, for example from Scenario 1 to 2, sharply reduces the amount of intermediate crop biofuel used. We find that intermediate crops are the cheapest compliance option to meet a GHG target or a renewable energy mandate, once the sub-mandates and caps have been complied with.

Study Scenarios
Fig 1. Energy consumption by fuel category by policy scenario

Problematic is the fact that the majority of intermediate crops globally are major commodity crops and their use in biofuel can be expected to cause indirect land use change (ILUC), just as with food-based biofuels. When we consider ILUC emissions, the very high total GHG emissions from intermediate crop soy biofuel significantly detract from the GHG savings of the policy as a whole. We can see this in Table 2, which shows the total GHG savings for each scenario, as well as the average cost of carbon abatement, the GHG credit price, and the total share of renewable energy in the road and aviation sectors.

Study Scenarios
Table 2. Environmental summary statistics

One important finding of this study is that a GHG target results in much greater GHG savings than a renewable energy mandate. Scenario 8, representing a 26% renewable energy mandate leads to a similar total amount of renewable fuel as Scenario 1, but delivers only around one-third the overall GHG savings. Consequently, the carbon abatement cost of Scenario 8 is around three times as high as that of Scenario 1. A GHG target also appears to be a much more cost effective means to achieve climate mitigation than a renewable energy mandate.

Renewable fuel policy is complex, and the impacts of policy changes are not always intuitive. Quantitative modeling, as demonstrated here, can be a useful tool in objectively analyzing a broad set of effects from changes in guidelines, and allows policy makers to make informed decisions.

The full report is available at https://theicct.org/publications/transport-carbon-intensity-targets-eu-aug2021.

About the ICCT

The International Council on Clean Transportation is an independent nonprofit organization founded to provide first-rate, unbiased research and technical and scientific analysis to environmental regulators. Their mission is to improve the environmental performance and energy efficiency of road, marine, and air transportation, in order to benefit public health and mitigate climate change.