Depending on their nature and severity, disasters can create large volumes of debris and waste. Waste volumes from a single event can be the equivalent of many times the annual waste generation rate of the affected community. These volumes can overwhelm existing solid waste management facilities and personnel. Mismanagement of disaster waste can affect both the response and long term recovery of a disaster affected area. Previous research into disaster waste management has been either context specific or event specific, making it difficult to transfer lessons from one disaster event to another. The aim of this research is to develop a systems understanding of disaster waste management and in turn develop context- and disaster-transferrable decision-making guidance for emergency and waste managers. To research this complex and multi-disciplinary problem, a multi-hazard, multi-context, multi-case study approach was adopted. The research focussed on five major disaster events: 2011 Christchurch earthquake, 2009 Victorian Bushfires, 2009 Samoan tsunami, 2009 L’Aquila earthquake and 2005 Hurricane Katrina. The first stage of the analysis involved the development of a set of ‘disaster & disaster waste’ impact indicators. The indicators demonstrate a method by which disaster managers, planners and researchers can simplify the very large spectra of possible disaster impacts, into some key decision-drivers which will likely influence post-disaster management requirements. The second stage of the research was to develop a set of criteria to represent the desirable environmental, economic, social and recovery effects of a successful disaster waste management system. These criteria were used to assess the effectiveness of the disaster waste management approaches for the case studies. The third stage of the research was the cross-case analysis. Six main elements of disaster waste management systems were identified and analysed. These were: strategic management, funding mechanisms, operational management, environmental and human health risk management, and legislation and regulation. Within each of these system elements, key decision-making guidance (linked to the ‘disaster & disaster waste’ indicators) and management principles were developed. The ‘disaster & disaster waste’ impact indicators, the effects assessment criteria and management principles have all been developed so that they can be practically applied to disaster waste management planning and response in the future.
Test results are presented for wall-diaphragm plate anchor connections that were axially loaded to rupture. These connection samples were extracted post-earthquake by sorting through the demolition debris from unreinforced masonry (URM) buildings damaged in the Christchurch earthquakes. Unfortunately the number of samples available for testing was small due to the difficulties associated with sample collection in an environment of continuing aftershocks and extensive demolition activity, when personal safety combined with commercial activity involving large demolition machinery were imperatives that inhibited more extensive sample collection for research purposes. Nevertheless, the presented data is expected to be of assistance to structural engineers undertaking seismic assessment of URM buildings that have existing wall-diaphragm anchor plate connections installed, where it may be necessary to estimate the capacity of the existing connection as an important parameter linked with determining the current seismic capacity of the building and therefore influencing the decision regarding whether supplementary connections should be installed.
This thesis is concerned with modelling rockfall parameters associated with cliff collapse debris and the resultant “ramp” that formed following the high peak ground acceleration (PGA) events of 22 February 2011 and 13 June 2011. The Christchurch suburb of Redcliffs, located at the base of the Port Hills on the northern side of Banks Peninsula, New Zealand, is comprised of Miocene-age volcanics with valley-floor infilling marine sediments. The area is dominated by basaltic lava flows of the Mt Pleasant Formation, which is a suite of rocks forming part of the Lyttelton Volcanic Group that were erupted 11.0-10.0Ma. Fresh exposure enabled the identification of a basaltic ignimbrite unit at the study site overlying an orange tuff unit that forms a marker horizon spanning the length of the field area. Prior to this thesis, basaltic ignimbrite on Banks Peninsula has not been recorded, so descriptions and interpretations of this unit are the first presented. Mapping of the cliff face by remote observation, and analysis of hand samples collected from the base of the debris slopes, has identified a very strong (>200MPa), columnar-jointed, welded unit, and a very weak (<5MPa), massive, so-called brecciated unit that together represent the end-member components of the basaltic ignimbrite. Geochemical analysis shows the welded unit is picrite basalt, and the brecciated unit is hawaiite, making both clearly distinguishable from the underlying trachyandesite tuff. RocFall™ 4.0 was used to model future rockfalls at Redcliffs. RocFall™ is a two-dimensional (2D), hybrid, probabilistic modelling programme for which topographical profile data is used to generate slope profiles. GNS Science collected the data used for slope profile input in March 2011. An initial sensitivity analysis proved the Terrestrial Laser Scan (TLS)-derived slope to be too detailed to show any results when the slope roughness parameter was tested. A simplified slope profile enabled slope roughness to be varied, however the resulting model did not correlate with field observations as well. By using slope profile data from March 2011, modelled rockfall behaviour has been calibrated with observed rockfall runout at Redcliffs in the 13 June 2011 event to create a more accurate rockfall model. The rockfall model was developed on a single slope profile (Section E), with the chosen model then applied to four other section lines (A-D) to test the accuracy of the model, and to assess future rockfall runout across a wider area. Results from Section Lines A, B, and E correlate very well with field observations, with <=5% runout exceeding the modelled slope, and maximum bounce height at the toe of the slope <=1m. This is considered to lie within observed limits given the expectation that talus slopes will act as a ramp on which modelled rocks travel further downslope. Section Lines C and D produced higher runout percentage values than the other three section lines (23% and 85% exceeding the base of the slope, respectively). Section D also has a much higher maximum bounce height at the toe of the slope (~8.0m above the slope compared to <=1.0m for the other four sections). Results from modelling of all sections shows the significance of the ratio between total cliff height (H) and horizontal slope distance (x), and of maximum drop height to the top of the talus (H*) and horizontal slope distance (x). H/x can be applied to the horizontal to vertical ratio (H:V) as used commonly to identify potential slope instability. Using the maximum value from modelling at Redcliffs, the future runout limit can be identified by applying a 1.4H:1V ratio to the remainder of the cliff face. Additionally, the H*/x parameter shows that when H*/x >=0.6, the percentage of rock runout passing the toe of the slope will exceed 5%. When H*/x >=0.75, the maximum bounce height at the toe of the slope can be far greater than when H*/x is below this threshold. Both of these parameters can be easily obtained, and can contribute valuable guideline data to inform future land-use planning decisions. This thesis project has demonstrated the applicability of a 2D probabilistic-based model (RocFall™ 4.0) to evaluate rockfall runout on the talus slope (or ramp) at the base of ~35-70m high cliff with a basaltic ignimbrite source. Limitations of the modelling programme have been identified, in particular difficulties with adjusting modelled roughness of the slope profile and the inability to consider fragmentation. The runout profile using RocFall™ has been successfully calibrated against actual profiles and some anomalous results have been identified.
