Published works

Earthquake hazard was not fully recognised in Australian building design until the mid-1990’s. This oversight has resulted in a legacy of vulnerable buildings that can be readily damaged in moderate to severe Australian earthquakes. In particular, older unreinforced masonry (URM) buildings built with the architectural styles, materials and construction details used in the United Kingdom are particularly vulnerable. Australian earthquakes have highlighted the vulnerability of this building type. These events include the Adelaide Earthquake of 1954, the Meckering Earthquake of 1968, the Newcastle Earthquake of 1989 and the Kalgoorlie Earthquake of 2010, all of which damaged pre WWII masonry buildings in particular.  Buildings of this style are present in the older centres of our major cities, and Melbourne has a very significant number of these. As shown in this research, in number nearly half of the buildings in the Melbourne central business district are of this type of construction. The damage to these buildings can greatly add to human casualties as a result of falling masonry elements. Further, the severity of damage and losses can impede the recovery of cities like Melbourne physically, economically and socially. Finally, many of these buildings have heritage value to communities that residents may want preserved.

This document reports on the final deliverable for Project A9 Cost-effective mitigation strategy development for building related earthquake risk of the Bushfire and Natural Hazards Collaborative Research Centre (CRC). It builds on the masonry component research of the University of Adelaide in this project and is a milestone for Geoscience Australia. The work follows the utilisation project entitled “Earthquake Mitigation of WA Regional Towns: York Case Study”, that was jointly delivered by GA and the University of Adelaide. The utilisation project developed original condition and mitigated vulnerability models for six URM building types. In this project, these outcomes have been applied to the much larger Melbourne CBD exposure.

The project had the following key components:

• Develop a building, business and demographic exposure database with the collected attributes tailored for modelling earthquake impact and the quantification of avoided consequences in economic terms.
• Attribute vulnerability models to the URM buildings in their present condition and with retrofit to a proportion after a thirty year program representing a future vulnerability condition.
• Simulate the damaging effects of a major earthquake on the present Melbourne CBD and at the end point of the retrofit program.
• Assess the change on scenario damage, casualties and economic losses as a result of the program.
• Assess the value of the program in terms of household willingness to pay to avoid losing heritage value buildings, drawing on UWA led research in the CRC.
• Assess the long term earthquake risk in the CBD precinct before and after the program of retrofit.

The work required the development of the three fundamental risk elements of earthquake hazard, community exposure and building vulnerability. It also entailed the assessment of the economic loss measures associated with human injury, contents losses, rental income, commercial property leasing, and business activity. Additionally, it included the application of the semi-intangible value placed on human life to society. Each of these are described below.

Earthquake hazard

This study has drawn upon the latest understanding of the Melbourne region earthquake hazard by utilising the recently released National Seismic Hazard Assessment (NSHA 2018) (Allen et al, 2018a). The bedrock hazard from this assessment shows Melbourne to have a “low” earthquake hazard by global standards but significant by Australian standards. The hazard is further amplified by the presence of the sediments deposited by the Yarra River.  These soil effects increase the hazard, particularly those in the study region south of the Yarra River. The effects of soil amplification can double the severity of shaking in some areas.

Community exposure

The definition of the building assets in the study region utilised several sources. The available state government building data integrated into the National Exposure Information System (NEXIS) was accessed and supplemented by an engineering survey database developed and maintained by GA for the Australian Reinsurance Pool Corporation. This was further refined by a desktop review of all masonry buildings utilising available street level imagery. In total there were 1,543 buildings in the study region, and 687 of these were identified as URM.

The assessment of human activity was achieved by utilising research undertaken outside of this project. This research utilised a population model developed by downscaling a destination zone based telecommunication model with pedestrian counts, the Melbourne traffic control systems movement counts, and building floor area information. Using this work it was possible to define the local human exposure at the time of the scenario event, particularly those in damaged buildings and those potentially exposed to falling masonry during a rapid onset earthquake event.

Building vulnerability

The building vulnerability assessment work for the URM building stock was a direct utilisation of the six vulnerability types identified in the earlier York WA mitigation study. This included the vulnerability in present condition, and that with mitigation measures applied to the vulnerable elements. To complete the context, the vulnerability of other building types was attributed using a suite of models developed through an adaptation of US HAZUS models, reference to  heuristically developed models from a GA facilitated UN workshop (Maqsood et al, 2014), and through heuristic adjustments by the project team. This vulnerability of non-URM buildings remained a constant in the study as mitigation of these buildings was not considered.

Economics of cost assessment

The economic assessment considered a broad range of measures. These ranged from the direct costs to property owner, building occupiers, and businesses through to health care costs and the partially intangible value placed on the loss of a human life. The aim was to provide scalable information on benefits versus cost to a range of decision makers and investors. Importantly, the measures where not comprehensive and so represent a lower bound to the actual avoided impacts mitigation achieves. For example, the cost of emergency response, clean-up and community recovery support were not considered. Neither was a macro-economic perspective developed to capture non-impacted businesses that would benefit from a stimulus in business activity such as in the construction industry, the supply of home appliances, soft furnishings and drapery. Significantly, the value of avoided heritage building loss was considered through the utilisation of metrics developed by a UWA led CRC project.

Scenario impacts and risks

The study considered a single rare earthquake scenario having an annual likelihood of 1/5,000 of causing the targeted bedrock shaking severity beneath the Melbourne CBD, or greater. This likelihoods corresponds with a 1%, chance of this shaking severity being exceeded in the next 50 years. For the event the injuries and other losses within the scope of this study were assessed using the human exposure corresponding with 11:00am of Monday through to Thursday. The losses ranged from $737m for building damage only, through to $1.66b for the other monetary costs considered.  The value of human life lost increased this to $3.97b. Where 25% of the masonry building stock was retrofitted, over 30 years, these losses reduced by approximately 16%.

The reduction in injuries if this event occurred in 30 years time was also evaluated. Serious injuries reduced by 16 and deaths by 98 persons. Urban Search and Rescue logistics would also reduce correspondingly.

In a similar manner, the long term financial risk of the Melbourne CBD study region was evaluated for building damage. It was presented as the average annualised loss for the URM building stock and for the entire study region buildings. It was also forecast 30 years into the future and the financial risk reduced by 38% for the URM building stock and by 10% across the entire study region buildings.

Discussion and outcomes

Earthquakes occur frequently in Australia with over 100 events greater than magnitude 3.0 (ML) recorded within the Australian continent every year by Geoscience Australia. The smaller and more frequent events are typically non-damaging, whereas the less frequent larger events can be very damaging when they occur close to a community. This plays out in the economics of strengthening older structures where the benefits of avoided building damage and contents losses through retrofit for earthquake are not a full offset for the significant costs.  Other avoided costs associated with business losses, lost wages, health care costs, and the value placed on human life, do increase the sum significantly but are not realised by the property owner. While not all avoided costs were considered, this project indicates that the justification for retrofit based solely on a financial investment may be difficult to demonstrate for URM buildings in Melbourne.

As was also illustrated by the earlier York study, there are other considerations for the retrofit of URM buildings in the Melbourne CBD and in other older business districts in the city. If a rare earthquake occurred locally during a period of high public exposure there would be considerable loss of life. This research has shown that if a 5,000 year Return Period (RP) event (5.5 Mw) occurring on a business day approximately 100 people would die with close parallels to the 2011 Christchurch Earthquake outcome for masonry structures (42 fatalities). This may point to cheaper levels of retrofit with the objective of tying back elements that could cause casualties, rather than having the aim of avoiding economic loss.

Further, following a rare, but credible, earthquake high value heritage buildings would be lost. The research has shown that the willingness to pay by just the residents of the City of Melbourne LGA adds a notional 10% of the total benefits of the mitigation program.

SUMMARY

The project has applied a range of retrofit measures for a suite of six URM building types developed as part of Project A9 to a very large population of URM building found in the Melbourne CBD. These measures have been demonstrated to reduce the physical vulnerability of each building. The project has also translated this vulnerability change into broader metrics that form an evidence base to inform decisions to retrofit.

The project has also demonstrated the benefit of retrofit through a virtual retrofit of a major city CBD. These benefits include reduced post event logistics for emergency management and the local government, reducing financial losses to building owners, businesses, and reducing injuries and fatalities. It has also demonstrated that retrofit reduces the long term financial cost of earthquake hazard, thereby making risk transfer through insurance uptake more affordable. Finally, it has demonstrated how valuable heritage structures can be progressively preserved for the future by protecting them from future credible earthquakes.