Configuring and Running


Words on this page that have been underlined and italicized can be found in the glossary on the right-hand side of the page.

During simulation, uncertainty is applied in the following order: floating (if applied), then duration ranges, then risk impacts.

With non-summary schedules, it is recommended to keep the number of iterations below 1,000 and use Latin Hypercube to achieve convergence quicker. A simulation with 1,500 activities and 1,000 iterations will take approximately 3 minutes minutes to complete (this may be slower or faster depending on your computer specs). Non-work modeling and floating will increase the time required during simulation. 

  1. Click the Simulations  driver.

  2. Click the New button in the Simulation Properties pane.

  3. Modify the parameters as needed.

  4. Click Run.

The Timestamp column shows the date and time the simulation completed in the UTC "timezone". Depending on daylight savings, this is typically 4 hours ahead of US Eastern Standard Time (EST) or 2 hours behind Central European Time (CET). 

Deleting Simulations

  1. Click the Simulations  driver.

  2. Select the desired simulation in the table

  3. Click the Delete button below the table.

As simulations are selected in the table, the Simulation Properties pane updates automatically.

Simulation Parameters

ParameterDefinition
NameA text field for summarizing the simulation.
MethodThe technique used for sampling. Whereas Monte Carlo uses repeated sampling, Latin Hypercube uses weighted sampling based on the number of iterations and the size of the range.
First BuildThe actual schedule used as the baseline for every simulation, which could be the deterministic schedule, but could also be the mean or mode, as calculated from the duration ranges of all the activities.
No. IterationsThe amount of times, or passes through the schedule, that sampling will take place.
SeedThe value that tells the random number generator where to begin, allowing the same sequence of random numbers to be generated over time. At the beginning of each simulation, the random number generator is seeded by converting the timestamp to a UNIX timestamp, and then it's stored for future reference.
Calculate Risk PriorityWith risk priority, the simulation is ran multiple times, each time removing the risk with the highest schedule or cost risk sensitivity, until no risks are left. For more info, see Risk Priority below.
Schedule DrivenAn option for choosing how risk priority will be calculated. When selected, risks with the highest schedule risk sensitivity are removed. If no schedule uncertainty has been included on the risk, ranging tabs, or non-work views, it will be disabled.
Cost DrivenAn option for choosing how risk priority will be calculated. When selected, risks with the highest schedule risk sensitivity are removed. If no cost uncertainty has been included on the risk or ranging views, it will be disabled.
Snap Project Start

An option for determining if the project start will be snapped to the start of the earliest object in the network.


For projects that were imported from XER files, this will always occur.

Snap Project Completion

An option for determining if the project completion will be snapped to the finish of the latest object.


When checked, the project completion snaps for the first build and then remains fixed for all subsequent iterations. For projects that were imported from XER files, this will always occur unless a planned completion date was entered.

Risk Priority

With risk priority, a simulation is ran multiple times, each time removing the risk with the highest impact from the prior run, until no risks are left. Highest impact is determined from the risk sensitivity, or the correlation between the probability of the risk and the project completion date. The sensitivity can be based on schedule or cost, depending on what was included for simulation and what the user decides.

The result is a tornado chart organizing risks by the potential number of days or dollars that could be gained if they were to be mitigated, from highest to lowest. For example: imagine a schedule with three risks, Risk A, Risk B, and Risk C, and schedule-driven. The simulation will proceed as follows:

RunIncluded Risks Risk with Highest Schedule Sensitivity
FirstRisk A, Risk B, Risk CRisk B
SecondRisk A, Risk CRisk A
ThirdRisk CRisk C
FourthNoneNone

Each run of the simulation is organized under a parent group in the Simulation tab. Individual runs cannot be deleted.

Any risk with a strategy of "Accept" in the risk register will NOT be removed during simulation, regardless of its sensitivity.

Sensitivity is calculated using the Spearman Rank Method. For more info, see Results and Reporting.

Only risks with a checkmark in the "Include in Simulation" column will take place during simulation.

Available Table Columns

For details on customizing columns, filtering, sorting, and organizing by WBS, see Interface Tour.

ColumnDefinition
NameA text field for summarizing the simulation.
TimestampThe date and time the simulation completed.
NotesA text field for adding notes and/or supplementary information.

PX Finish Date

The "X" percentile finish date (p value), i.e. finish date value observed during at least X% of iterations. P values available from P5 to P95 in the Finish Date submenu.


Jump to:


Still Need Help?

Click the icon above or visit our support desk.


 Glossary

Deterministic: an outcome or scenario that is determined or chosen always with the same result, for example for a duration, date, schedule, or other value.

Latin Hypercube: A technique that iterates through the schedule many times, choosing inputs based on weighted sampling (based on the number of iterations and the size of the range) to calculate the probability of different outcomes such as completion dates. A Latin Hypercube simulation may require up to 30% fewer iterations for the simulation to converge than with the traditional Monte Carlo method. 

Monte Carlo: A technique that iterates through the schedule many times, choosing inputs based on repeated random sampling from a cumulative probability distribution, to calculate the probability of different outcomes such as completion dates. A Monte Carlo simulation may require a large number of iterations to ensure that the values have been sampled from the entire range.

P value: The value (date, duration, etc.) which was observed during at least X% of iterations. For example, the P80 Finish Date is the finish date observed during at least 80% of the iterations. To put it another way, 80% of the iterations completed on that date or earlier.

Risk Sensitivity: The correlation between risk impacts and the duration of the project.


Next Steps