Schedule Risk Analysis with Monte Carlo Simulation and Completion Probability

Learn how Monte Carlo simulation turns task estimates into a completion date probability, how to read P50 and P80 results, and how to size contingency.

3 min read2 viewsSchedule Risk Analysis with Monte Carlo Simulation and Completion Probability
— · 0 votes
Contents
  1. How a Monte Carlo schedule simulation works
  2. Reading P50 and P80 on the S-curve
  3. Finding risk drivers and sizing contingency
  4. Questions about Monte Carlo schedule risk
  5. What is schedule risk analysis?
  6. How many iterations does a Monte Carlo schedule simulation need?
  7. Why is the Monte Carlo finish date later than the plan?
  8. Do you need special software for Monte Carlo schedule risk analysis?

Schedule risk analysis with Monte Carlo simulation estimates the chance a project finishes by a chosen date. Instead of one fixed duration per task, each task gets a range of possible durations. A computer then recalculates the schedule many times, drawing a random duration for every task. The results form a probability distribution of completion dates, giving a project completion probability estimate for any deadline.

How a Monte Carlo schedule simulation works

A Monte Carlo schedule simulation starts from a normal network schedule with task links and durations. Each uncertain task receives a duration distribution built from a three-point estimate of optimistic, likely and pessimistic values. Triangular and beta distributions are two shapes that describe how durations spread around the likely value. Every iteration recalculates the critical path, the longest chain of dependent tasks, and records its finish date.

Reading P50 and P80 on the S-curve

The S-curve of completion probability plots each possible finish date against the chance of finishing by then. The P50 date is the point the project meets or beats in half of the simulated runs. The P80 date is met in eight of every ten runs, so it sits later on the curve. Choosing a confidence level depends on the cost of lateness and the owner's appetite for risk.

Finding risk drivers and sizing contingency

Sensitivity analysis shows which tasks drive the spread of finish dates in a probabilistic schedule risk model. A tornado chart ranks those tasks as horizontal bars, with the strongest driver placed at the top. The schedule contingency reserve is the time gap between the single-estimate finish and the chosen confidence date. Narrowing the ranges of leading drivers through better planning can shrink the reserve the project needs.

Questions about Monte Carlo schedule risk

What is schedule risk analysis?

Schedule risk analysis is a method that measures how uncertainty in tasks affects the project finish date. The analysis replaces single duration estimates with ranges and identified risk events. The output is a range of possible finish dates, each with a probability attached. Teams use the result to set realistic deadlines and decide where to focus mitigation.

How many iterations does a Monte Carlo schedule simulation need?

The number of iterations a Monte Carlo schedule simulation needs depends on schedule size and the precision wanted. More iterations make the S-curve smoother and the P80 date more stable between runs. A practical check is to rerun the model and compare the key dates. When repeated runs give nearly the same P50 and P80, the iteration count is sufficient.

Why is the Monte Carlo finish date later than the plan?

The Monte Carlo finish date often lands later than the plan because of merge bias in the network. Merge bias means a task waits for the latest of several preceding paths to finish. When each path can run late, the chance that at least one does grows. Skewed duration ranges, where possible overruns exceed possible savings, push the curve later as well.

Do you need special software for Monte Carlo schedule risk analysis?

Yes, Monte Carlo schedule risk analysis needs a tool that samples random durations and recalculates the schedule. Dedicated schedule risk programs read the task network directly from planning software. A spreadsheet with random number functions can handle a small network with simple links. The quality of the result depends more on honest duration ranges than on the tool.

Read next

Latest articles