Bottom-Up Modeling of Building Stock Dynamics - Investigating the Effect of Policy and Decisions on the Distribution of Energy and Climate Impacts in Building Stocks over Time

Doctoral thesis, 2019

In Europe, residential and commercial buildings are directly and indirectly responsible for approximately 30–40% of the overall energy demand and emitted greenhouse gas (GHG) emissions. A large share of these buildings was erected before minimum energy-efficiency standards were implemented and are therefore not energy- or carbon-efficient. Consequently, buildings offer significant potential in terms of energy efficiency and the reduction of GHG emissions compared to the status quo. To make use of this potential at scale, targeted policy measures and strategies are needed that should be based on a quantitative assessment of the feasibility and impact of these measures.
Building stock models (BSMs) have long been used to assess the current and future energy demand and GHG emissions of building stocks. Most common BSMs characterize the building stock through the use of archetype buildings, which are taken to be representative of large segments of the stock. The increasing availability of disaggregated datasets—such as building registries, 3D city models, and energy performance certificates—has given rise to building-specific BSMs focusing on describing the status quo as an input to energy planning, primarily on the urban scale. Owing to the availability of building-level data, BSMs can be extended beyond policy advice and urban planning, to the assessment of large building portfolios. Thus far, the advances made in building-specific BSMs on the urban scale have not been transferred to the national scale, where such datasets are often not available. Moreover, the focus on an increasingly detailed description of the existing stock has left approaches for modeling stock dynamics without much development. Stock dynamics, therefore, are still primarily modeled through exogenously defined retrofit, demolition, and new construction rates. This limits the applicability and reliability of model results, as the influence of economic, environmental, or policy factors on stock development is not considered.
This thesis addresses these shortcomings and advances modeling practices in BSMs. The thesis with appended papers provides a methodology for how the modeling of national building stock can be further developed in terms of building stock characterization through synthetic building stocks as well as stock dynamics through the use of agent-based modeling. Furthermore, the thesis extends BSM applications to inform the strategic planning of large building portfolios through the integration of a maintenance and renovation scheduling method to project the future development of building portfolios.