Send us Fan Mail

A quantitative risk assessment puts a probability, a cost impact, and a schedule impact on each quantified risk. A Monte Carlo simulation then samples the full set thousands of times. The result is a range of outcomes. You fund your total risk exposure, not individual risks.

Nate Habermeyer and Albert Brier, Director, Project Controls at Dokainish, break down how a QRA works. Albert uses two six-sided dice to visualize probability. They then cover the difference between qualitative and quantitative scoring, the risk register versus the risk universe, the alternatives to Monte Carlo simulation, and Albert's advice for anyone starting to plan a capital project, since most projects dramatically undervalue risk on day one. Teams typically run a QRA before a funding or pre-execution gate, so the inputs need to be right before the model runs.

Albert uses two six-sided dice to visualize probability. They give 11 possible totals, yet no total carries a one-in-11 chance. One combination makes a two, and seven is the most likely total. Add more dice and the results grow complicated. Albert says Dokainish does a QRA by rolling many dice together, one for each risk and estimating range. Monte Carlo simulation does this thousands of times and shows where the results concentrate.

A risk register covers only the risks you named. Albert calls the full set of credible threats and opportunities the risk universe. Systemic risk comes from how your organization operates, such as a board that meets monthly with an agenda set three months ahead. Base uncertainty is the range around an estimate before any discrete risk applies. A full accounting of quantitative impact includes both.

The first simulation gives an aggregate model, and funding decisions use it. Project managers need the disaggregated view, which sensitivity runs produce by including and excluding each risk. Albert explains why you fund total exposure and not individual risks. He also describes the risk workshop, which can take a week on a register of a few hundred risks, and the three inputs the model needs, a schedule, a cost estimate, and a risk register.

Albert describes parametric estimating, which draws on historical project data, such as a large company's database of tens of thousands of capital projects. He points listeners to Bent Flyvbjerg's writing on reference class forecasting. He also covers the limits of probability times impact. That method gives the mean but no range, and it handles schedule poorly. He shares his view that the register as a large spreadsheet will probably give way to a network of causes, events, consequences, and responses.

Teams usually run a QRA before a funding or pre-execution gate. At that point the inputs can still change, and the project team is about to carry the register into execution. The workshop before the gate sets the quality of the model and of the register that follows.

Some teams treat a completed register as the full risk universe. A full accounting adds base uncertainty and systemic risk. The workshop before the gate is a strong opportunity to correct the inputs.

Read the companion blog post: https://dokainish.com/insights/what-is-a-qra/

Presented by Dokainish & Company www.dokainish.com

The Risky Planner podcast delivers expert insights on project controls, capital project management, and strategic planning for today's complex business environment. Subscribe for regular episodes featuring industry leaders and practical advice.