Risk – can we predict the unpredictable? Should we?

19 July 2016
5 min

Infrastructure underpins national economies. Businesses, industries and communities all intrinsically rely on power, water or telecommunications networks as their lifeblood for operation. They all rely on road and rail networks to facilitate the movement of goods and people – movement in our modern day trading world is as good as money.

Most businesses and communities take these underlying systems for granted. They will always be there, won’t they? Is ‘always’ a dangerous assumption? Tasmania, an idyllic island which is home to over 500 000 people, might answer: Yes.

Over the last century, Tasmania has been largely reliant on its own hydroelectrical generation. When Basslink, an undersea cable connecting the Tasmanian grid to the Australian mainland, failed on 22 December 2015, it led to a power outage from which Tasmania still hasn’t recovered. At the same time, Tasmania was suffering from a major drought, which meant that its capacity for generating hydroelectrical power was greatly diminished.

Tasmania’s engineering systems, on which the entire State relied, would have been designed for a certain level of resilience. In engineering speak: a 1:100 year event. But what happens when these rare events happen cumulatively? One of these ‘one in a hundred year’ events puts a massive strain on an area’s infrastructure and systems, but two simultaneous and interconnected events is disastrous. In their case, their systems came to a grinding halt.

Our obligation is not only to develop design inputs that create better contingency, but also have the potential to stay one step ahead of an uncertain future.

Rethinking “this will never happen to us”

A more pertinent question to pose might be: are these major events happening more frequently as a result of human influence? And if they are, do businesses need to rethink their “this will never happen to us” plans?

Tasmania is not alone… The January 2011 Brisbane floods were some of the most devastating in Australian history: beginning with bad state-wide flooding in December, the situation escalated dramatically when a monsoon arrived on Christmas Eve from the Coral Sea, culminating in the Brisbane River’s banks bursting on 11 January. This rapid sequence of disastrous events caused 35 fatalities, affected over 200 000 people throughout Queensland and caused damage of around AUD2.4 billion.

As both natural hazards and human-induced disasters occur more and more frequently, it’s imperative that we get better at harnessing emerging digital technologies to foresee and fend them off. Is a risk management strategy possible for the unpredictable nature of disasters – and if it is, how might we integrate it into both our operations and designs?

Making peace with uncertainty

Modern day risk analyst, statistician and essayist Nassim Nicholas Taleb, believes that: In today’s evolving world, uncertainty is on the rise with increasing severity. Because rare events are impossible to predict, he argues that we need to make peace with this uncertainty and to reduce our negative exposure.

What would this acceptance mean to business owners, governments, insurance agencies and the general public? Should we recalibrate our design criteria on the basis that the world is shifting and ‘one in a hundred year events’ are now happening once a decade?

Architecting risk buoyancy

We must strive for a world in which unknowns can be identified, assessed, evaluated and this information then utilised to improve the end product. Managing risk is about leveraging predictive capabilities that extend to the unknown, so that uncertainty is replaced with expectancy. This, in essence, is called ‘risk buoyancy’.

Designing for the unexpected

Imagine a world where buildings can regularly withstand the blows of a big-scale disaster, because the design accounted for a typical one in a hundred year event to wreak havoc every decade. Although not, as of yet, commonplace, more and more examples are in evidence:

San Francisco's multibillion dollar Seismic Retrofit Program allows the region’s lifeline infrastructure and housing stock to rebound more quickly from future seismic activity. Five years after the earthquake, Christchurch has concentrated on rebuilding a resilient city focused on low damage design which can, quite literally, ‘bounce back’ faster after any major incident.

Japan has similarly prepared for future earthquakes; New Orleans' USD14.5 billion network of levees, floodwalls and pumps guarantee to ward off destruction from another event similar to Hurricane Katrina; Copenhagen is building parks that turn into ponds during extreme rainfall events.

In our journey of predicting the unpredictable, we must seek out these examples of robust or anti-fragile architecting as examples of positive risk mitigation. Future risk-based thinking needs to consider the multiscalar and multidimensional attributes that, in combination, create uncertainty.

In an ever-changing world, radically underscored by uncertainty: Is it possible to eliminate all unpredictability by determining our unknowns? One thing we can be sure of is that we will never know the answer. But we do know that we want to create an environment where unforeseen factors can be integrated into the creative process to provide end users with an asset that is risk buoyant.

Perhaps we should change our traditional design approach to consider the 1:100 year event to be a dice with a hundred sides that is thrown every year. Every time, there is a 1 per cent chance of the event occurring. Every time, it reminds us there is a risk of it happening more than once in a hundred years. That way, not even a ‘one in a hundred year event’ will catch us off guard. We would already have anticipated it and, more importantly, put in place mechanisms of thriving beyond its potential devastation.


This blog was authored by: Simon Van Wyk



Unfortunately, you are using a web browser that Aurecon does not support.

Please change your browser to one of the options below to improve your experience.

Supported browsers: