## Monte Carlo simulations using event chain methodology

This forum includes discussion about project risk analysis and risk management theory: Monte Carlo simulations, Event Chain Methodology, schedule and cost risk analysis. Please submit questions and case studies about your experience with our project risk analysis software.

Moderator: Intaver Support

PMBOK
Posts: 10
Joined: Wed Jan 17, 2007 1:51 pm
Location: Pittsburg, PA, PPG

### Monte Carlo simulations using event chain methodology

Could you explain the process of simulations based on event chain methodology? What is the difference between classic Monte Carlo simulations as it defined in PMBOK and simulations using ecm?

Intaver Support
Posts: 921
Joined: Wed Nov 09, 2005 9:55 am
Once events, event chains, and event subscriptions are defined, Monte Carlo analysis of the project schedule can be performed to quantify the cumulative impact of the events. Probabilities and impacts of events are used as an input data for analysis.

In most real life projects, even if all the possible risks are defined, there are always some uncertainties or fluctuations in duration and cost. To take these fluctuations into account, distributions related to activity duration, start time, cost, and other parameters should be defined in addition to the list of events. These statistical distributions must not have the same root cause as the defined events, as this will cause a double-count of the project’s risk.

Monte Carlo simulation process for Event chain methodology has a number of specific features. Before the sampling process starts all event chains should be identified. Particularly, all sender and receiver events should be identified through an analysis of state tables for each activity. Also, if events are assigned to resources, they need to be reassigned to activities based on resource usage for each particular activity. For example, if manager is equally involved in two activities, a risk “Manager is not familiar with technology” with a probability 6% will be transferred to both activities with probability of 3% for each activity. Events assigned to summary activities will be assigned to each activity in the group. Events assigned to lags are treated the same way as activities.

Each trial of the Monte Carlo simulation includes the following steps specific to Event chain methodology:

1. Moments of events are calculated based of statistical distribution for moment of event on each state.

2. Determines if sender events have actually occurred at this particular trial based on probability of the sender.

3. Determines if probabilities of receiver events are updated based on sender event. For example, if a sender event unconditionally causes a receiver event, probability of a receiver event will equal 100%.

4. Determines if receiver events have actually occurred; if a
receiver event is a sender event at the same time, the process of determining probabilities of receiver events will continue.

5. The process will repeat for all ground and excited states for all activities and lags.

6. If an event that causes the cancellation of an activity occurs, this activity will be identified as canceled and the activity’s duration and cost will be adjusted.

7. If an event that causes the start of another activity occurs, such as execution of mitigation plan, the project schedule will be updated for the particular trial. Number of trials where the particular activity is started will be counted.

8. The cumulative impact of the all events on the activity’s duration and cost will be augmented by accounting for fluctuations of duration and cost.

The results of the analysis are similar to the results of classic Monte Carlo simulations of project schedules. These results include statistical distributions for duration, cost, and success rate of the complete project and each activity or group of activities. Success rates are calculated based on the number of simulations where the event “Cancel activity” or “Cancel group of activities” occurred. Probabilistic and conditional branching, calculating the chance that project will be completed before deadline, probabilistic cashflow and other types of analysis are performed in the same manner as with a classic Monte Carlo analysis of the project schedules. Probability of activity existence is calculated based to two types inputs: probabilistic and conditional branching and number of trials where an activity is executed as a result of a “Start activity” event.
Intaver Support Team
Intaver Institute Inc.
Home of Project Risk Management and Project Risk Analysis software RiskyProject
www.intaver.com