Complexity Risk and Modeling Disorder

Complexity Risk and Modeling Disorder

John K. Hollmann (Validation Estimating, LLC, USA)
DOI: 10.4018/978-1-5225-1790-0.ch007


Despite 50 years of cost accuracy research, companies are generally unable to quantify the worst outcomes. In the process industries about 10 percent of large projects overrun their budgets by 70 percent or more. The system behavior of these blowouts often reflects disorder. For complex projects, the blowout proportion is 15 to 30 percent of projects. Many risk analysts ignore the worst outcomes as “unknown-unknowns” or “black swans”; but they are neither—we know the causes and their impact is somewhat predictable. Cost disasters start with a mix of systemic weakness and risk events. The cost of mundane projects may overrun by 20 to 40% which is bad but no disaster (financiers assume they will overrun by 25%). Add complexity and stress and the projects can cross a “tipping point” into disorder and chaos with cost overruns of 50, 100 or 200 percent-true disasters. This chapter describes complexity risk and the disorder it can lead to, practical measures of complexity and stress and how to incorporate those measures in non-linear risk quantification models.
Chapter Preview

Project Cost Behavior

My focus is on large projects involving construction; this may include:

  • Process, power and utility plants,

  • Mines,

  • Civic and industrial infrastructure, and even

  • Complex buildings.

For these and related industry sectors, cost estimate accuracy (i.e., actual outcome/funding estimate) is well documented and dismal. In the actual case, 10 percent of projects overrun by 70 percent or more after normalization for scope change and price escalation. (Hollmann, 2012). The average project overruns its’ sanctioned amount by about 20 to 25% (as stated, smart bankers assume this). Industry’s failure to realistically quantify nominal risks (e.g., overruns <50%) is addressed in Chapter 3 on systemic risks and in my book “Project Risk Quantification” (Hollmann, 2016); this chapter is focused on what causes and how to forecast the long tail at >50% overrun; the realm of complexity and blowouts.

Complete Chapter List

Search this Book: