This thesis presents the application of data science techniques, especially machine learning, for the development of seismic damage and loss prediction models for residential buildings. Current post-earthquake building damage evaluation forms are developed for a particular country in mind. The lack of consistency hinders the comparison of building damage between different regions. A new paper form has been developed to address the need for a global universal methodology for post-earthquake building damage assessment. The form was successfully trialled in the street ‘La Morena’ in Mexico City following the 2017 Puebla earthquake. Aside from developing a framework for better input data for performance based earthquake engineering, this project also extended current techniques to derive insights from post-earthquake observations. Machine learning (ML) was applied to seismic damage data of residential buildings in Mexico City following the 2017 Puebla earthquake and in Christchurch following the 2010-2011 Canterbury earthquake sequence (CES). The experience showcased that it is readily possible to develop empirical data only driven models that can successfully identify key damage drivers and hidden underlying correlations without prior engineering knowledge. With adequate maintenance, such models have the potential to be rapidly and easily updated to allow improved damage and loss prediction accuracy and greater ability for models to be generalised. For ML models developed for the key events of the CES, the model trained using data from the 22 February 2011 event generalised the best for loss prediction. This is thought to be because of the large number of instances available for this event and the relatively limited class imbalance between the categories of the target attribute. For the CES, ML highlighted the importance of peak ground acceleration (PGA), building age, building size, liquefaction occurrence, and soil conditions as main factors which affected the losses in residential buildings in Christchurch. ML also highlighted the influence of liquefaction on the buildings losses related to the 22 February 2011 event. Further to the ML model development, the application of post-hoc methodologies was shown to be an effective way to derive insights for ML algorithms that are not intrinsically interpretable. Overall, these provide a basis for the development of ‘greybox’ ML models.
“One of the most basic and fundamental questions in urban master planning and building regulations is ‘how to secure common access to sun, light and fresh air?” (Stromann-Andersen & Sattrup, 2011). Daylighting and natural ventilation can have significant benefits in office buildings. Both of these ‘passive’ strategies have been found to reduce artificial lighting and air-conditioning energy consumption by as much as 80% (Ministry for the Environment, 2008); (Brager, et al., 2007). Access to daylight and fresh air can also be credited with improved occupant comfort and health, which can lead to a reduction of employee absenteeism and an increase of productivity (Sustainability Victoria, 2008). In the rebuild of Christchurch central city, following the earthquakes of 2010 and 2011, Cantabrians have expressed a desire for a low-rise, sustainable city, with open spaces and high performance buildings (Christchurch City Council, 2011). With over 80% of the central city being demolished, a unique opportunity to readdress urban form and create a city that provides all buildings with access to daylight and fresh air exists. But a major barrier to wide-spread adoption of passive buildings in New Zealand is their dependence on void space to deliver daylight and fresh air – void space which could otherwise be valuable built floor space. Currently, urban planning regulations in Christchurch prioritize density, allowing and even encouraging low performance compact buildings. Considering this issue of density, this thesis aimed to determine which urban form and building design changes would have the greatest effect on building performance in Central City Christchurch. The research proposed and parametrically tested modifications of the current compact urban form model, as well as passive building design elements. Proposed changes were assessed in three areas: energy consumption, indoor comfort and density. Three computer programs were used: EnergyPlus was the primary tool, simulating energy consumption and thermal comfort. Radiance/Daysim was used to provide robust daylighting calculations and analysis. UrbaWind enabled detailed consideration of the urban wind environment for reliable natural ventilation predictions. Results found that, through a porous urban form and utilization of daylight and fresh air via simple windows, energy consumption could be reduced as much as 50% in buildings. With automatic modulation of windows and lighting, thermal and visual comfort could be maintained naturally for the majority of the occupied year. Separation of buildings by as little as 2m enabled significant energy improvements while having only minimal impact on individual property and city densities. Findings indicated that with minor alterations to current urban planning laws, all buildings could have common access to daylight and fresh air, enabling them to operate naturally, increasing energy efficiency and resilience.
