Primavera P6 Schedule Risk Analysis: Guide | Vose Software

Primavera P6 Schedule Risk Analysis

From a deterministic plan to finish dates you can commit to Last updated August 2026

What is schedule risk analysis?

Schedule risk analysis applies Monte Carlo simulation to a project schedule. Task durations become ranges instead of single values, identified risk events are added with their probability and impact, and the simulation runs the schedule thousands of times — producing distributions of finish dates and cost rather than one date that everyone quietly disbelieves. The project can then commit at a stated confidence level, such as a P80 finish date.

The starting point is the plan you already maintain: a Primavera P6 or Microsoft Project schedule with its network of tasks, links and calendars. The analysis does not replace that plan — it measures how much the plan's single answer should be trusted, which risks threaten it most, and what date and budget are realistic to promise.

Why deterministic schedules are optimistic

Projects very commonly cost more and take longer than planned, and the bias is structural rather than a matter of poor planning. A deterministic schedule carries one duration per task — usually close to the most likely value. But task durations are right-skewed: there are many more ways for work to take longer than expected than to finish early, so the realistic average already sits above the most likely value, task by task.

The network then amplifies the bias. Wherever parallel paths converge, the merge point must wait for the latest incoming path — so even if each path is only occasionally late, the chance that none of them is late shrinks with every extra path. Deterministic dates ignore this merge effect entirely; simulation prices it automatically, which is why simulated finish dates are almost always later than the deterministic one.

The third amplifier is correlation. Task durations do not vary independently: the same weather, the same crew productivity, the same design maturity drive many tasks at once. When durations move together, the good and bad luck stop cancelling out, and the spread of possible finish dates widens — a schedule risk analysis that ignores correlation will look reassuringly precise and be wrong.

How do you perform a schedule risk analysis?

  1. Import and audit the schedule. Bring the Primavera P6 or Microsoft Project schedule into the risk tool and check it is fit for simulation: complete logic links, sensible constraints, no dangling tasks. A schedule that only works because of hard-coded dates cannot respond to simulated variation.
  2. Describe everyday uncertainty. For tasks with meaningful variability, define uncertainty on the work — how much there is, and how productively it will be executed — rather than guessing percentages task by task. Grouping similar activities into work types keeps the assessment consistent and fast.
  3. Add risk events from the risk register. Discrete threats and opportunities — a delayed permit, late material delivery, rework, a disruption that halts part of the project — enter with a probability and an impact, distinct from everyday variability.
  4. Let common factors drive correlation. Model shared drivers (productivity factors, weather, approvals) as factors that affect groups of tasks together, so correlation emerges from causes instead of an arbitrary correlation matrix.
  5. Simulate and read the results. Run the simulation and read the distributions: the P50 and P80 finish dates, the drivers ranked by their effect on the outcome, and the cost consequences of schedule slip.

Steps 2–4 are where judgement lives, and they are also where a risk workshop earns its time: the ranges and risks should come from the people who own the work, calibrated against how similar projects actually went — not from a habit of ±10%.

What to model beyond duration ranges

The most common failure of schedule risk analyses is stopping at step 2 — putting three-point ranges on durations and simulating. That captures everyday variability but misses the things that actually blow schedules up. A credible model distinguishes several mechanisms, because they behave differently and are managed differently:

Work amount vs productivity. A task can take longer because there turned out to be more work (scope), or because the work went slower (productivity). Productivity factors typically affect whole families of tasks at once — which is precisely what makes them dangerous, and what makes them a natural source of correlation. Risk events come in shapes: an event delaying a specific task; identical independent events hitting a set of similar tasks; an event that changes productivity across many tasks; extra-work risks (probabilistic branching — a design fails review and a rework loop appears); and disruption events that temporarily halt a whole section of the project.

Cost belongs in the same model. Schedule slip drives cost through time-dependent costs (site overheads, standing charges), and risk events carry direct costs of their own. Simulating time and cost together yields the joint picture — including the awkward outcomes where the project is late and expensive — that separate analyses miss. For the budgeting side of that picture, see how to calculate cost contingency.

Reading the results: P-dates, drivers and the stochastic Gantt

The headline outputs are the percentile dates. A P80 finish date is the date the project finishes on or before in 80% of samples — a commitment with a stated one-in-five chance of being missed. The gap between the deterministic date and the P80 date is the schedule contingency, exactly parallel to the P80 logic of cost contingency. Histogram and cumulative plots of finish date, duration and cost make the whole distribution visible, not just one number — below, Tamara's duration histogram for a construction schedule, with every percentile readable in the statistics panel and two risk scenarios compared side by side:

Tamara simulation results: duration histogram of a commercial construction schedule comparing managed and unmanaged risk scenarios, with a percentile statistics table

The management value is in the drivers. Tornado-style sensitivity charts rank the uncertainties and risk events by their influence on cost and time, turning the simulation into a prioritised action list; re-running after a mitigation shows exactly how many days and euros it buys. A stochastic Gantt chart shows the likely range of each task's start and finish — planning honesty at task level — while a scatter plot of finish date against cost reveals the joint risk picture, and cashflow projections over time put a realistic band around funding needs.

Tamara finish-date tornado chart ranking construction tasks by their influence on the project finish date

Tamara stochastic Gantt chart showing probabilistic start and finish ranges for construction tasks, with P5 to P95 dates for a selected task

Schedule risk analysis for P6 and MS Project with Tamara

Tamara, our schedule risk analysis tool, implements this workflow directly on Primavera P6 and Microsoft Project schedules. Import performs a schedule health check — scoring the schedule's quality and flagging the specific issues that would undermine a simulation — and when the master schedule changes, one click re-imports it with the risk model intact, so the analysis lives with the plan instead of being a one-off exercise. Below, the audit report Tamara produces while importing a schedule — Monte Carlo suitability and quality checks scored against goals, with the offending tasks listed:

Tamara schedule audit on import: Monte Carlo suitability and quality checks with red and green scores against goals, and the list of tasks with excessive float

Uncertainty is then described the way the previous sections set out: work amount and productivity uncertainty by work type, the full family of risk-event shapes including probabilistic branching and disruptions, cost uncertainty alongside time, and correlation induced naturally through shared productivity factors.

Scale is not a constraint: Tamara has been tested on real plans of up to 50,000 tasks, returns results for typical 50–300-task schedules in a fraction of a second, and runs 5,000 samples of a 34,000-task project in under ten minutes — fast enough to explore scenarios in a live workshop, switching risks on and off to compare mitigation strategies. Outputs include every chart described above — percentile dates, tornado charts, stochastic Gantt, cost-vs-finish scatter, cashflow projections — all copyable straight into Word or PowerPoint reports, plus an integrated spreadsheet for costs and risks that live outside the master schedule.

Tamara Desktop costs €2,150 per user per year — see the price list — and the free trial imports your own P6 or MS Project schedule, so the first analysis can be on the project you are actually running. For how Tamara compares with other tools in this space, see our review of the top 6 project risk analysis tools.

Frequently asked questions

What is schedule risk analysis?

Monte Carlo simulation applied to a project schedule: duration ranges and risk events replace single-point estimates, and the simulation produces distributions of finish dates and cost so commitments can be made at a stated confidence level such as P80.

What does a P80 finish date mean?

The date the project finishes on or before in 80% of simulation samples — a commitment with roughly a one-in-five chance of being missed. The gap between the deterministic date and the P80 date is the schedule contingency.

Can I run schedule risk analysis on a Primavera P6 schedule?

Yes — Tamara imports Primavera P6 and Microsoft Project schedules directly, health-checks them during import, and re-imports the updated master schedule in one click with the risk model intact.

Why model risk events separately from duration ranges?

Ranges describe variability of work that will certainly happen; risk events happen only sometimes. Folding events into ranges hides their all-or-nothing effect on the schedule's tail — and hides them from the mitigation list.

How long does a schedule risk simulation take?

With Tamara's engine it is not a bottleneck: a fraction of a second for typical 50–300-task schedules, and 5,000 samples of a 34,000-task project in under ten minutes.

Tamara logo

Tamara

Schedule risk analysis for your projects

Import your Primavera P6 or MS Project schedule, add uncertainty and risks from your register, and commit to dates and budgets at a confidence level you choose.