The energy demand sectors industry, transport and buildings are together directly responsible for around 51 % of the global energy-related CO2emissions and indirectly drive the emissions in the energy supply sectors. The demand sectors are characterized by many subsectors, technologies, heterogeneous end-users with different preferences and needs, and rapid changes. This complexity is not easy to capture in global, long-term models analysing climate change mitigation pathways. Adding more detail does not neccesarily improve the capability of models, that have such a broad scope to analyse long-term future change. Therefore, these models typically contain a more stylized representation of energy demand futures. The consequence of this is that also much less attention has been paid to the use of energy and the role of energy reduction in a global setting to achieve climate targets. In this thesis, it is analysed whether and how the energy demand dynamics representation in global model assessing long-term climate change can be improved.
The first part of the thesis studies the current representation of energy demand futures in integrated assessment models (IAMs) by relating model assumptions, structure and input to outputs and evaluating the model projections to historic data and sector specific studies. Over the coming decades, if current trends remain unchanged, global energy demand is projected to continue to grow. This is despite improved energy efficiency. In 2050 projected global energy demand in respectively the buildings, industry and transport sector ranges from 180-220, 190-240 and 160-190 EJ. Scenario analysis shows that in order to meet internationally agreed climate targets both energy efficiency and fuel switching (including electrification) play an important role, with fuel switching being dominant in the long run. Specifically for the transport sector, trends are generally comparable to historical indicators of activity growth, modal shift, energy intensity, income and price elasticities. Fuel switching trends however go beyond historical measurements. Models pay relatively little attention to changes in the sectors activity to mitigate emissions, which could potentially complement, other strategies.
Better understanding and describing of the development of future energy demand could allow for greater policy relevance. We show for the electric vehicle transition case that new insights can be attained by dynamically modelling complex demand-side processes in a transparent manner while keeping the model relatively simple. A key challenge ahead is to, where possible, identify simple cause-effect relationships based on empirical data to better understand energy demand response. Moreover, using multiple scenarios, comparing different model types and performing sensitivity analysis on uncertain assumptions are important tools to improve our understanding of and exploring the full solution space energy demand response. Key avenues for future research are analysing behavioral change, the physical requirements, such as infrastructure and material demand underlying the energy demand projections, alternative demand sector climate policy, possible barriers to sectoral transitions and the relation between short term and long term as well as local and global opportunities.