Share on Facebook Tweet about this on Twitter Share on LinkedIn
Why Project Risk Analysis Software?

Project risk analysis software helps you to determine how risks and uncertainties can affect project schedule, rank risks, rank projects based on risk exposure, determine crucial tasks, and perform other types and analysis. What is project risk analysis software and how can you decide which version to use in your organization can be difficult. There are some very solutions available, all offering various capabilities and different price points. The question becomes what are the most important features required for running project risk analysis and what are secondary functionalities that you might consider.

Monte Carlo Simulation

The first thing to recognize is that all activities are characterized by uncertainties. These uncertainties are variances in start times, finish times, duration, cost etc. that can occur in any activity due to risk events or natural variances. This means that regardless of how much effort you put into your planning, the execution of the project will be probabilistic as activities start or end later than planned, costs fluctuate due to variances in performance, or risks could occur. Because projects are probabilistic, you need a software that can account for these probabilities and currently the best method we know to measure the impact of these uncertainties Monte Carlo simulation. If the software does not support Monte Carlo (or the related Latin Hypecube), then you should remove it from your list.

However, just have Monte Carlo simulation is not enough. Any quality Monte Carlo simulation software will include a variety of statistical distribution types which are used to model uncertainties or risk impacts as probability density functions. Look for a variety of distribution types include the common Triangular, Lognormal, Normal as well as others such as Beta, or BetaPert. On the other hand, for project risk analysis, having 30 different statistical distributions is not required as generally the ones mentioned here will suffice 99.9% of the time.

Performance is also important, but less so than in the past as Moore’s Law has provided extremely powerful computers, such that even basic laptops are able to perform risk analysis on very large schedules in a reasonable time. However, if your schedules are large ( > 1000 activities, resource loaded, or have many risks events (drivers), you should consider that Addins that use the Excel calculation engine suffer as it is not optimized to run Monte Carlo simulations at this scale.

Risk Events vs Uncertainties in Project Risk Analysis Software

Project activities can be affected by two different types of risk: risk events and uncertainties. Risk events, also referred to as risk drivers or discrete risks, can be defined using probability of occurrence and the severity of impact or consequence. Risk events can cause sudden shifts in the state of activities or projects. In addition, risk events can be managed and consequences reduceded through planning. Uncertainties are defined by statistical distributions and are used to describe the variance that is inherent to all activities. Unlike risk events, uncertainties are not manageable and are irreducible.

Project risk analysis software should be able to model both type of risk . Typically, risk events are added and modeled using an integrated risk register that allows you to assign the risk to activities or resources with probability and impacts on cost and schedule. Typically, uncertainties are modeled my providing low and high estimates and a statistical distribution for duration, but more advanced software will allow distributions to be assigned to Start Times, Finish Times, Lags, Cost, and Income.

Integrated Cost and Schedule Project Risk Analysis

Project risk analysis software should also support cost and/or schedule analysis. In addition, you should be able to run both at the same time “integrated cost and schedule risk analysis”. Along with project schedules, the software must support both fixed and resource costs. The value of this type of analysis is that no only do you generate risk adjusted cost and schedule estimates, but it will also take into account time dependent costs that are derived from the variance in work done by resources assigned to activities.

Integration of Project Risk Analysis Software with Other Scheduling Software

Most project management teams, especially those involved, will have a variety of tools used as part of the estimating, planning, and management process, this includes scheduling and cost estimating software. At a very minimum, project risk analysis software should be able to import schedules developed in the most popular scheduling software such as Microsoft Project, Primavera, or Asta PowerProject. In addition, project risk analysis software must be able to accurately process project actuals so the current project performance is correctly factored into the results of the Monte Carlo Analysis.

It is not only important to bring data into the software. You should also be able to move or export the results of your simulations as either risk adjusted schedules or raw data to the other tools that your team is currently using. Whether this is an explicit export function or a simple cut and paste, make sure that your results can used with other applications as necessary.

Results of Project Risk Analysis and Reports

There are 3 basic analysis and visualizations that are part of Monte Carlo simulation: Probability Histograms, Cumulative Probability, and Sensitivity Analysis. The software should be able to provide you this analysis both at the Project level and for every activity in the schedule. The probability histograms and cumulative probability plots should provide results for Duration, Finish Times, and Cost. You should be able to customize these charts to provide quick feedback on confidence levels and other pertinent details.

Sensitivity analysis provides information on which activities have the potential to cause the most variance to project objectives such as Cost, Finish Time, and Duration. The most popular method for calculating sensitivity is the Spearman Rank Correlation Coefficient.

Reporting is also extremely important and the software should have the capability to generate reports that include all simulation data and visualizations for 1 or more activities. Project risk analysis is often focused on deliverables or other interim milestones and ability to report multiple activities is very convenient.

Nice to have features

In addition to the core functionality outlined above, most software packages will offer additional features that add value to the project risk analysis.
Scenario analysis or multiple baselines: this allows you to model and compare how different risk management strategies, mitigations, and responses impact project objectives and differences in pre and post mitigation risk scores.

Cost and Cash Flow Analysis: views that provide specific analysis for the results of cost, or the combination of cost and income (cash flow) with NPV discounting.

Conditional and Probabilistic Branching: models alternative project plans based on branching nodes. Provides expected values for your project where there are multiple possible paths (branches) that could occur based on either certain conditions or probabilities.

Factors or Issues: These are techniques that share characteristics of both risk events and uncertainties and are used to model common factors that can generate uncertainty on project objectives. Like risks they can have impacts that are assigned to activities and resources with impacts, but have 100% probability. Factors or issues allow you break out sources of uncertainty like variances in performance or production.

Weather Calendars: allows you to create ranges of weather related non-working days that can be based on weather records for specific locations.
Schedule Quality Analysis: schedule risk analysis is only valid if performed on high quality schedules. At minimum, the software should check for circular logic, hard constraints, and dangling activities (no predecessor or successors).

Master Schedules: Large schedules are often broken down into subprojects, that are reassembled for periodic analysis.
Convergence Monitoring: simulation results converge when they results of the simulation do not change significantly (as defined by a % change in the Mean and Standard Deviation over x iterations). This was a popular feature years ago when running a simulation could take a many hours. While not as critical, it can be useful for times when you want to run a quick but valide analysis on large projects.

Lastly, project risk analysis software needn’t break your budget. When this type of software was first introduced it could command a premium price to provide the capability needed in very large and complex projects. This is no longer the case and there are high quality software packages available at a fraction of the price that was seen even a decade ago. If, as it turns out, you are thinking of performing project risk analysis on your projects, hopefully this will quickly allow you to create a shortlist of potential software packages that can meet your needs. Once you have a list, most vendors will provide you a free evaluation period which should allow you to to understand how each package fits into your workflow and if it is a good match for you.