Schedule Risk Analysis (Monte Carlo) Basics for Planners

What Is Schedule Risk Analysis?

Schedule risk analysis (SRA) is a quantitative method that evaluates the likelihood of completing a project on time. Unlike a deterministic CPM schedule that shows a single finish date, SRA uses probability distributions to model uncertainty in activity durations and the impact of risk events. The most common technique is Monte Carlo simulation, which runs thousands of schedule iterations to produce a range of possible completion dates and their probabilities.

For planners, SRA answers questions like: What is the probability of finishing by the contractual deadline? or Which activities drive schedule risk?

Duration Uncertainty vs. Risk Events

SRA models two main types of uncertainty:

• Duration uncertainty: Each activity’s duration is not fixed but follows a probability distribution (e.g., triangular, PERT-beta). You define optimistic, most likely, and pessimistic durations based on historical data or expert judgment.
• Risk events: Discrete events that may or may not occur (e.g., permit delays, weather shutdowns). Each risk event has a probability of occurrence and an impact on duration or cost. These are often modeled as “risk branches” in the schedule.

The simulation randomly samples from these distributions and risk branches for each iteration, building a histogram of finish dates.

How Monte Carlo Simulation Works

In a typical Monte Carlo run:

1. The scheduler defines duration ranges for all or key activities.
2. Risk events are added with probabilities and impacts.
3. The software runs 1,000–10,000 iterations. In each iteration, it randomly picks a duration for each activity (from its distribution) and determines whether risk events occur.
4. The finish date for each iteration is recorded.
5. After all iterations, a cumulative probability curve (S-curve) is generated, showing the probability of finishing by any given date.

Most scheduling tools (Primavera P6 Professional, Oracle Primavera Risk Analysis, @RISK) support Monte Carlo simulation. For a quick check, you can also use Project Assure, a free browser-based tool that parses your XER locally and runs DCMA 14-point checks, EVM, and schedule risk metrics—no data upload required.

Reading a P80/P50 Result

The simulation output often includes percentiles:

• P50 (median): There is a 50% probability of finishing on or before this date. This is the “most likely” finish date if you consider uncertainty.
• P80: There is an 80% probability of finishing on or before this date. Often used as a contingency date.
• P90 or P95: Used for high-confidence targets.

For example, if your deterministic schedule says “Finish: 1 June” but the P80 is 15 June, you have only about a 20% chance of meeting the deterministic date. The difference between P50 and P80 (e.g., 10 days) is the schedule contingency needed to achieve 80% confidence.

When SRA Is Worth Doing

Schedule risk analysis is not needed for every project. Consider SRA when:

• The project has high complexity or many interdependent activities.
• There are tight contractual deadlines with penalties.
• Significant uncertainty exists in durations (e.g., R&D, first-of-a-kind projects).
• Stakeholders require quantitative confidence levels (e.g., government, lenders).
• You need to justify contingency time or budget.

SRA is less useful for small, repetitive projects with low uncertainty or when the schedule is poorly built (e.g., missing logic, hard constraints). Always validate the deterministic schedule first.

Common Pitfalls

• Garbage in, garbage out: If duration ranges are guessed without data, results are meaningless.
• Ignoring correlation: Activities that share resources or weather should be correlated, or the simulation may underestimate risk.
• Over-reliance on P50: P50 is not a “target”; it’s a median. Use P80 or higher for commitments.
• Forgetting risk events: Duration-only SRA misses discrete risks like strikes or design changes.

Practical Steps for Planners

1. Build a logically sound, resource-loaded schedule in Primavera P6.
2. Identify critical and near-critical paths.
3. Assign three-point estimates (optimistic, most likely, pessimistic) to key activities. Use historical data or expert judgment.
4. List risk events with probability and impact. Add them as separate activities with “if-then” logic or use risk registers.
5. Run Monte Carlo simulation (e.g., 1,000 iterations).
6. Analyze the output: probability curve, sensitivity chart (tornado diagram), and criticality index.
7. Present results to stakeholders: “We have an 80% chance of finishing by 15 June, requiring 10 days of contingency.”

Schedule risk analysis turns guesswork into data-driven decisions. By understanding P50/P80 and the drivers of uncertainty, you can set realistic deadlines, allocate contingency wisely, and communicate confidence levels clearly. For a lightweight way to start exploring schedule risk, try Project Assure—it runs entirely in your browser, so your XER never leaves your machine.