A Global and Local Learning Model of Transport (GALLM-T)
MODSIM · 2017
The transport sector is accounts for 14% of global greenhouse gas emissions (Sims, et al., 2014) but is also an essential service underpinning economic growth and societal well-being.The transport sector will therefore need to maintain or enhance mobility while transitioning to lower emission modes, fuels and technologies.Globally the road sector is the largest source (72%) of transport emissions.Most road vehicle fuel efficiency improvements reduce the total cost of travel and therefore represent a negative abatement cost.However, other abatement opportunities such as vehicle electrification and other alternative fuels involve switching vehicle technologies, with emerging technologies initially having higher costs than the existing more emissions-intensive alternative.Uptake of these technologies helps reduce their costs through 'learning by doing', where cost reduces as uptake increases.Economies of scale in manufacturing of these vehicles will also reduce their cost.To explore the impacts of policy and economic drivers on the transport mix, CSIRO has developed a partial equilibrium model of the global transport sector, GALLM-T, which explicitly includes learning by doing.The model uses experience curves to endogenously determine the future cost and uptake of fuel conversion technologies related to transport.GALLM-T features 13 regions, 17 fuel conversion technologies, 16 fuels, 5 passenger modes and 7 modes of freight transport.Technologies subject to learning include batteries in electric vehicles and fuel cells in fuel cell vehicles.Component learning is included where technology components have shared learning among technologies that have those components.For example, the carbon capture and storage (CCS) component in fuel production facilities is a component that is shared among all facilities that include CCS technologies.The nonlinear experience curves have been approximated as piecewise-linear functions, and the model's `selection' of which linear piece it is on at each point in time forms the core integer part of GALLM-T.GALLM-T is solved in GAMS as a mixed integer linear program.This paper introduces the model and provides results from its application in the second of CSIRO's National Outlook projections, a project which explores different futures for Australia, within the global context, through quantitative scenario analysis.In the Outlook, global transport, electricity, land use and economic models are linked to generate a consistent set of inputs for national models that explore Australian outcomes in more detail.Future demand for transport has been sourced from linking GALLM-T with a global general equilibrium model (GTAP.ME-3).GALLM-T has also been linked with CSIRO's GALLM-E model, which is used to determine the future cost and uptake of electricity generation technologies.Both GALLM-E and GALLM-T have the capacity to project uptake of batteries and fuel cells.Where this occurs in both models, the combined impact pushes these technologies down the experience curves faster, accelerating the rate of cost reduction.This paper compares results from scenarios with moderate and strong global climate action.Illustrative results show a more than 60% share of electric drive trains in the total stock of passenger vehicles and 70% in light commercial vehicles by 2050 under both carbon price scenarios ($31/tCO2e and $65/tCO 2 e by 2050).There is a limited uptake of fuel cell drive trains in cars and light commercial vehicles.Low coal prices also lead to the construction of coal to liquids plants (with and without CCS), mainly providing fuel for the freight sector which continues to have a high share of diesel engines.There is a greater share of production of biofuels, which displaces conventional and alternative fossil fuels under the higher carbon price scenario from the year 2040 onwards.