Reevaluating Risk Modelling in Australian Construction: Have We Been Doing It Wrong?.

The persistent occurrence of ‘unexpected cost blowouts’ on large-scale Australian infrastructure projects suggests that our risk management approach is just plain wrong.

Risk analysis is a cornerstone of effective planning and execution. Traditionally, two primary techniques are employed: qualitative assessments and quantitative risk assessments. While qualitative assessments categorise risks as High, Medium, or Low, quantitative assessments express risks in numerical terms, combining potential loss amounts with probabilities or frequencies. However, the conventional approaches to measuring and allocating risk may be fundamentally flawed, particularly when considering the nature and scale of potential risks. Furthermore, based on the general ‘100% certainty’ of large-scale project blowouts, it’s evident that this traditional process has failed.

This article argues that adopting the power law distribution, rather than relying on the normal distribution, may provide a more accurate framework for understanding and mitigating risks in construction projects.

The Conventional Approach and Its Limitations

Consider the quantitative assessment of risk in a typical scenario. As a basic example, the risk of budget blowouts on a small construction project might be estimated with a potential overrun ranging from $1 million to $10 million and an annual frequency of 0.5. This results in an annualised loss expectation of $0.5 million to $5 million. At first glance, this seems like a straightforward calculation.

However, when examining real-world data, the limitations of this approach become evident. For instance, for example, the following is a list of large Australian projects and their respective initial and final costs:

  • Sydney Opera House
    • Original Budget: $7 million AUD
    • Final Cost: $102 million AUD
    • Percentage Overrun: Approximately 1,357%
  • Parramatta Light Rail
    • Original Budget: $1.6 billion AUD
    • Revised Cost: $2.4 billion AUD
    • Percentage Overrun: Approximately 50%
  • Adelaide Desalination Plant
    • Original Budget: $1.1 billion AUD
    • Final Cost: $2.2 billion AUD
    • Percentage Overrun: Approximately 100%
  • Melbourne’s East West Link
    • Original Budget: $6 billion AUD
    • Estimated Cost: $10 billion AUD (before project cancellation)
    • Percentage Overrun: Approximately 67%
  • Sydney Light Rail
    • Original Budget: $1.6 billion AUD
    • Final Cost: $2.9 billion AUD
    • Percentage Overrun: Approximately 81%
  • NSW Government Schools Project
    • Original Budget: $5.3 billion AUD
    • Revised Cost: $6.7 billion AUD
    • Percentage Overrun: Approximately 26%
  • Wonthaggi Desalination Plant, Victoria
    • Original Budget: $3.1 billion AUD
    • Final Cost: $4 billion AUD
    • Percentage Overrun: Approximately 29%
  • Perth Children’s Hospital
    • Original Budget: $1.2 billion AUD
    • Final Cost: $1.6 billion AUD
    • Percentage Overrun: Approximately 33%
  • Snowy 2.0
    • Original Budget: $2 billion AUD
    • Revised Cost: $12 billion AUD
    • Percentage Overrun: Approximately 500%

Introducing the Power Law Distribution

The discrepancies observed in risk assessments may be better understood through the lens of the power law distribution. Unlike the normal distribution, the power law distribution acknowledges that extreme events, though less frequent, have a disproportionately significant impact. This concept is prevalent in various natural and social phenomena, such as the frequency and magnitude of earthquakes. For every earthquake measured at 1 on the Richter scale, there is one at magnitude 2 for every ten at magnitude 1, and one at magnitude 3 for every hundred at magnitude 1, and so on.

Applying the power law to construction risk analysis suggests that severe budget blowouts, while less likely, are far more impactful than the normal distribution would predict. This perspective shifts our understanding of risk from expecting a narrow range of outcomes to anticipating a broader spectrum where catastrophic events are plausible.

Practical Implications for Construction Risk Management

To incorporate the power law distribution into construction risk management, we may need to move beyond traditional risk assessment models and adopt a scenario-based approach. This involves presenting least, median, and worst-case scenarios based on an order-of-magnitude difference between them, evaluated over an extended period, such as ten years. This timeframe allows for significant changes within the business and project environments, resulting in evolving risk profiles.

Implementing a Power Law-Based Strategy

  1. Least-Case Scenario: This represents the most frequent, lower-end risks that, while not devastating individually, can collectively impact project success. These risks should be managed through routine risk treatment measures.
  2. Median-Case Scenario: This encompasses moderate risks that occur less frequently but still pose substantial threats. Organizations should prepare contingency plans for these risks, focusing on risk acceptance and proactive mitigation.
  3. Worst-Case Scenario: This involves the rare but highly consequential risks. Planning for these scenarios requires robust risk-avoidance strategies, comprehensive insurance coverage, and strategic reserves to manage potential financial impacts.

Qualitative Assessment Translation

While the power law distribution may seem quantitative, its principles can enhance qualitative risk assessments as well. Clear communication about the existence of worst-case scenarios and the inclusion of “black swan” events in risk analyses is crucial. This approach fosters a more nuanced understanding of risk categories, enabling better-preparedness and strategic decision-making.

Conclusion

The potential to adopt the power law distribution approach in construction project risk analysis would be a significant shift. By recognising the potential for extreme, high-impact events and planning accordingly, project managers can develop more resilient and adaptable risk management strategies. This shift not only enhances the accuracy of risk assessments but also ensures that organizations are better equipped to handle the full spectrum of potential challenges, ultimately leading to more successful project outcomes. One thing is certain, our current approach is just plain wrong.

Author: Peter Bowden – Principal, Wildara Australia Pty Limited.