EVM and Project Risk Analysis

A question that we are often asked is how does project risk analysis (PRA) compare or relate to earned value management (EVM)? It often goes further, if I am doing EVM, is there a need for risk analysis, or is it redundant? In this case, while there is a little overlap, risk analysis and earned value are complementary practices as part of an integrated project/program management process. This begs the question, how do the two complement each other. First, let’s define earned value and project risk analysis.

Project risk analysis is a process of identifying project risk events and uncertainties and assessing their expected impacts on key project objectives such as cost, schedule, and performance using Monte Carlo simulations. EVM was developed to allow project teams to measure how well they are executing their project by comparing planned work against actual work performed over predetermined intervals. Earned value takes into account measures of planned work such as Budgeted Cost of Work Scheduled  (BCWS)  vs Budgeted Cost of Work Performed (BCWP), which provide an indicator of how much work was actually performed vs what was planned. Depending on the particular system, there are a plethora of such metrics which can seem very complex, but they answer a simple, yet very important question: Has the project produced the value for the funds provided? This lies at the heart of how contracts are managed as the payment is based not upon how much work or effort has been done by the project, but by the value that has been created. If you know that you are only going to get paid based upon delivering value according to plan, then it is incumbent upon you to ensure that your plan is realistic. This is where project risk analysis supports the goals of the EVM process.

One of the key objectives of project risk analysis is to create “risk adjusted” realistic project plans. It does this by assessing the possible impacts of risks and uncertainties on project objectives, minimizing the impact of these risks through mitigation or response planning, and adding appropriate schedule margin and cost contingency to account for unmanageable risk. This process provides a provides a high confidence that the project can be executed according to plan. As, we can see, initially project risk analysis is performed as part of the development of the project plan; however, project risk analysis can also augment the EVM process during project execution as part of the forecasting of delivery dates.

EVM is used not only to track actual performance, but can include indicators that use current performance to estimate time to complete or cost at completion. These indicators ETC (Estimate to Complete) or EAC (Estimate at Completion) provide values that forecast the required budget to complete the project (ETC= Estimated cost for uncompleted work)  or how much the entire project will cost (EAC = sunk costs + ETC) One of the main issues with these indicators is that they are single point estimates, which are notoriously inaccurate when estimating under conditions of uncertainty (risks and uncertainties). Project risk analysis can augment the EVM performance tracking by providing forecasts that account for future risks and uncertainties. In this scenario, project plans be analyzed to predict possible outcomes given both the actual project performance and the risks and uncertainties that project has yet to encounter. This is extremely useful as this provides a risk adjusted forecast given current project status.

 

In the example above, we can see a plot that shows the project schedule plan vs. actuals with a forecast that shows how the actuals will affect the possible schedule outcomes.  In the EVM world, the analysis would provide duration to calculate an EAC for a schedule as shown by the central blue line. But in reality, there is range of possible outcomes that can be expressed using confidence levels or percentiles (P values).  In the example above, we can see while the EAC is calculated based on a single date, there is range that can be achieved from the very aggressive (P10) to very conservative and more likely (P80). In this case, the P80 represents a finish date at that is much more feasible than the one provided by the deterministic calculation. In practical terms, this means that risk analysis will help to determine statistical distributions of remaining project finish time and remaining cost. These distributions will be used to calculate EAC based on selected percentile. This process cnn be repeated for different project milestones will taking into an account actual project performance.

So we can see that when used in concert, EVM and project risk analysis can e improve not only the chance that projects will deliver the planned value on time and budget, but can also be used together to understand how current project performance may impact future project execution.