Fifty Years On from Cyclone Tracy: Galvanizing Progress for Australia Wind-Resilient Building Design
Peter DatinDecember 23, 2024
Image: Town Hall Ruins, Darwin
December 25, 2024, marks the fiftieth anniversary of Cyclone Tracy’s devastation of the Northern Territory city of Darwin, Australia. The event was a watershed moment for cyclone resiliency and wind-resilient structural design in Australia, with significant enhancements implemented to construction standards and building codes in cyclonic wind environments from the lessons learned.
Since 1974, further wind code and construction practices were made with insights from Tracy and subsequent cyclones to impact Australia.
However, Tracy’s legacy on cyclone risk management, our understanding of building vulnerability to cyclone winds and construction to resist the impact of winds, lives on.
Impact of Cyclone Tracy
In the early hours of Christmas Day 1974, Cyclone Tracy tracked slowly, directly over the city of Darwin (Fig. 1). A small, intense cyclone, Tracy’s recorded winds reached a strong category 3 on the five-point Australian Bureau of Meteorology scale before the anemometer at Darwin Airport failed at 3:10 a.m. on December 25, 1974, a full fifty minutes before the storm’s eye passed overhead.
Satellite and damage observations indicated the storm’s strength was a category 4 (sustained 10-minute wind of 159-198 kilometers per hour (99-123 miles per hour); and gust wind speeds of 226-280 kilometers per hour (140-174 miles per hour)).
Figure 1: Cyclone Tracy’s path over northern Australia, December 1974
Darwin had a population of approximately 48,000 in 1974, around half of the Northern Territory population. At the time, residential construction was predominantly timber frame with asbestos cement wall sheeting (also then known as ‘fibro’).
Tracy caused immense structural damage, primarily from wind and predominantly to residential structures. Sixty-six people were killed and around 60 percent of homes in Darwin were destroyed, while in the northern suburbs, this figure approached 100 percent.
Only six percent of residences were livable in the storm’s immediate aftermath.
Changes to Australian Wind Loading Codes
Wind loading building codes were in place in Australia in 1974, which did account for the wind loads exerted from a cyclone of the strength of Tracy, though typically these were applied to commercial (and industrial) construction and not residential, leaving the residential building stock vulnerable.
There had been earlier cyclones including Cyclone Althea in 1971 in Townville, Queensland where lessons had been implemented into new mid-70s residential construction.
However, in the newer northern suburbs of Darwin, engineering assumptions in the codes had not accounted for the horizontal loads imparted on the structures – i.e., only the vertical loads on the roofs were included in the building design. More generally, the then-early versions of wind codes had limitations highlighted in Cyclone Tracy.
A subsequent investigation and findings post-Tracy resulted in notable learnings and improvements to the Australian wind code: namely, fatigue testing of metal roofing under cyclical wind loading, the inclusion of internal wind pressure in the design (accounting for openings such as windows in structures broken due to windborne debris), and the inclusion of horizontal wind pressures in the design.
During the 1970s and 1980s, various updates to wind loading standards occurred as research commenced to understand what caused the buildings in Darwin to perform so poorly. New editions of the Australian wind loading standard (AS 1170.2) first introduced in 1973, saw either updates or amendments in 1975, 1978, 1981, and 1983.
A major revision to AS 1170.2 in 1989 first introduced the concept of ultimate limit state design, likely the first wind design standard in the world to do so.
In 2002, the first version of the standard was published in a single, combined document for Australia and New Zealand (AS/NZS 1170.2:2002) which included many major changes.
Another edition followed in 2011 which included, among other updates, a windborne debris design requirement. The latest version, published in 2021, incorporates additional modifications to enhance the wind design of buildings and for the first time introduces a ‘climate change multiplier’ for cyclonic regions.
Most recently, a 2024 amendment to the wind loading standards saw lessons from Cyclone Seroja (2021) incorporated – including updates to regional design wind speeds, internal pressure provisions, and loading parameters for roof-mounted solar panels.
Figure 2: Wind design regions in Australia (AS/NZS 1170.2:2021, Amendment No. 2 – 2024)
The benefits from these improvements in wind design code specifications that resulted from Cyclone Tracy have been demonstrated in recent cyclones including Cyclones Larry (2006) and Yasi (2011).
Most of the damage to residential buildings in these storms was suffered by houses constructed before 1980, while those built to modern codes, incorporating the lessons learned from Cyclone Tracy, suffered far less damage.
Thoughts on the Future
It should be remembered that approximately 40-50 percent of houses in cyclone-prone areas of Australia are pre-1980 houses. The current building codes generally focus on new building designs and do not directly address these older structures.
Targeted retrofits on older houses in cyclone-prone areas could lead to significant damage reductions for future events. They should be considered a priority for building code committees and officials and local and regional governments.
It is also worth noting that as with Tracy, testing the performance of newer properties in cyclone winds has seen failures – damage investigations by the Cyclone Testing Station following Cyclone Debbie in 2017 revealed significant damage to modern residential properties from wind-driven water ingress.
My Moody’s RMS colleagues and I attended the twenty-second Australasian Wind Engineering Society Workshop in Townsville, Queensland. Cyclone Tracy was a focus for this anniversary year, as it has been in many communities across Australia.
One of the dominant themes of the workshop was community disaster resilience to tropical cyclones. Building codes since Cyclone Tracy have focused on the performance of individual buildings; however, community disaster resilience is better defined as no loss (or minimal loss) of community functionality (e.g., key services including first responders, basic infrastructure including power, performance of all buildings including pre-1980 buildings, etc.).
Discussions about community disaster resilience will continue into the foreseeable future and will likely lead to further refinements in building codes and a reduction in damages from future storms.
Regardless, as the fiftieth anniversary of the devastation on Christmas Day 1974 approaches, we can acknowledge the impact of Cyclone Tracy today on building codes in cyclone-affected regions of Australia.
Conclusion
From the catastrophe modeling perspective, Moody’s RMS Australia Cyclone Model vulnerability module references the Australia wind code design requirements and their changes over time, reflecting the vulnerability of building stock to cyclone winds.
The learnings from Cyclone Tracy and subsequent cyclones in the updated wind design standards are referenced in our current cyclone model and will continue to be referenced in our new cyclone model under development.
Peter leads the climate vulnerability group for Moody's RMS models, overseeing wind and convective storm vulnerability development in North America and the Caribbean, Australia, Europe, and the Asia-Pacific regions.
He has been the lead vulnerability developer for the North Atlantic Hurricane and the North America Severe Convective Storm models for over 10 years. In addition to vulnerability development, Peter has led multiple field reconnaissance teams to assess hurricane and tornado damage in the U.S. and Caribbean as well as research trips to Australia and other regions to understand local building practices.
Prior to joining RMS (now Moody's) in 2011, Peter received his Ph.D. in Civil Engineering (structural/wind engineering) from the University of Florida in Gainesville, where he researched structural load paths in wood-frame structures and worked on structural testing under simulated wind loads of various building components and materials.
Peter also holds a master’s degree in Civil Engineering from Clemson University and a bachelor’s degree in Civil Engineering from Oklahoma State University. He has also worked as a part-time lecturer at Stanford University.
