@RISK Enables Powerful Project Management for Edmonton Light Rail Transit System

Customer

SMA Consulting

Industry

Construction & Engineering, Consulting & Legal, Logistics & Transportation

Product(s)

@RISK & TopRank
Z

Application

Project Management

A Growing Metropolis

The Canadian city of Edmonton, Alberta, is poised for big things. Thanks to an abundance of research institutions and the oil and gas reserves in its surrounding regions, this city is a center of education, technology and economic growth. According to the 2011 Canadian census, the area had the second-fastest pace of population growth of any metropolitan area in Canada, and most of this was in the 25-44 age range, indicating a bright future in terms of income and consumer spending for this northern metropolis.

In an effort to accommodate this momentum, the City of Edmonton decided to expand their public transportation, known as the Light Rail Transit (LRT) system, from just one line to a network that connects all sectors of the city. While the ultimate goal of a comprehensive LRT system is not expected to be completed until 2040, major steps have already been taken, including a 3.3 km extension to Edmonton’s north LRT system. Of course, projects like these are expensive—anywhere from $200-$800 million—and come with considerable uncertainty. To deal with this, the City brought on a team to help them hone in on the costs and risks associated with the project, with a goal of greater cost and schedule certainty. As part of that team, SMA Consulting’s Risk and Project Controls Manager Jesse Kostelyk was tasked with estimating the potential costs and analyzing the uncertainties involved in the project. To do so, he turned to @RISK.

Relying on @RISK @RISK allows for the project management modeling that Kostelyk and SMA Consulting need, enabling them to estimate the impact of various risks on the cost and schedule of the project. Thanks to @RISK’s integration of Microsoft Project schedules with Microsoft Excel models, the consulting team could manipulate schedule inputs and then observe how those changes would affect the budgetary outcome of this important, multi-million dollar project. The result was a much more objective, precise view of the scope of the LRT expansion for decision-makers.

By leveraging the spreadsheet environment as an interface to a project schedule, @RISK enables modeling flexibility that was never possible in Microsoft Project alone. “@RISK’s ability to integrate models in Excel and Project opens up a world of possibilities for capturing and quantifying construction project uncertainty,” reports Kostelyk. “At SMA Consulting, it’s our go-to program for risk analysis.”

Abounding Uncertainties

Kostelyk and his team needed to give Edmonton decision makers a clear understanding of what the project would entail. “Our goal was to inform the decision makers of the range of costs and schedule end dates that could be realized on this project to allow them to plan with confidence,” says Kostelyk.

This goal was challenging, as any construction project is rife with a variety of unknowns: weather variability, fluctuations in worker productivity, changes in material costs, or broad shifts in the overall market. With all of these uncertainties at play, how can project managers pull it all together to determine interrelationships between cost, schedule and risk? Traditionally, project planners estimate a single, static point for the risk level on each uncertainty—but this does not accurately portray reality.

Summary

The City of Edmonton, Alberta, needed to determine costs and risks associated with expanding the Light Rail Transit system, with a goal of greater cost and schedule certainty. SMA Consulting used @RISK to estimate potential costs and analyze uncertainties involved in the project.

Categorizing Risks

SMA’s team took a more nuanced approach, looking to create a distribution of possibilities to demonstrate the true likelihood of certain risks materializing and their probable impacts. To do this, they categorized the different types of uncertainty and evaluated them separately.

According to Kostelyk, there are three primary areas of uncertainty that they needed to model: general uncertainty, escalation and schedule uncertainty, and risk factor uncertainty. The first category, general uncertainty, is included in all aspects of a construction project. “Every line item on an estimate has some uncertainty to it; it’s intrinsic to the process,” says Kostelyk. “Some things have more uncertainty than others—bolts are pretty easy to estimate, but not the amount of soil the contractor will have to move in order to install a foundation.” All of these typical factors and their associated risks are lumped into the general uncertainty category.

“What the general uncertainty category doesn’t include are specific risk events, such as river flooding or environmental contamination, that may or may not happen” says Kostelyk. “If something like that happens, there may be an impact on many different items simultaneously.” Thus, the ‘risk factor uncertainty’ category was created to lump these types of risks together for analysis.

The ‘escalation uncertainty’ category accounts for the fact that the overall market and general economy will change over time, and that it is impossible to fully predict both what the market will be doing when a project begins and the uncertainty related to the project schedule. Typically, inflation is assumed.

Kostelyk explains: “In construction planning, estimators may just slap on 5% per year for escalation on the whole project. So a project that takes two years would have an overall escalation of 10%.” However, this one-size-fits-all estimation can fall widely off the mark. “For this project, we took the approach of acknowledging that this amount could be more or less, and taking into account the fact that there is more uncertainty on when the different parts of the project will be built, when these market increases or decreases will happen and what the actual economics will be.”

Nevertheless, the real crux of the problem, according to Kostelyk, is how to define and then model the uncertainty in each category for the project. SMA used a heuristic approach to assigning general uncertainty distributions to each cost item. “We identified the factors of general uncertainty such as level of design detail and flexibility of innovation and then assigned a high, medium, low scoring.” The result of this approach allowed customized Beta General distributions to be developed in @RISK.

SMA then went through numerous risk assessment workshops to identify risk events to the project as well as the probability they would occur and impact to each cost and schedule item. They made use of their context-based risk identification approach and, again, used custom heuristics for turning probabilities and impacts into distributions. “SMA is linked closely with university research on defining what people mean when they say something is ‘likely’ to happen, or ‘unlikely’, etcetera. We then turn these descriptors into binomial distributions in @RISK.”

To model escalation uncertainty, Kostelyk and his team separated the project components into different escalation types. Local roadwork, for example, has a different escalation profile than the signal systems and cabling. Next, they gathered historic and projected escalation values from various sources, and determined the range in each year for each type. At this stage, they integrated the @RISK capabilities with MS Project to model the project schedule and duration-related uncertainty.

“When we put all this together into an integrated model in @RISK, we call it the black box,” says Kostelyk.

Balancing Cents and Certainty

After running the risk analysis using Monte Carlo simulation, the project managers were able to bring their results to the Edmonton City Council and present them with a defined cost range. “They are asked how much risk are you willing to take to move this project forward,” says Kostelyk, “and then we can quantify how certain we are about how much it will cost.” For example, for this project, if the City wanted a specific lower-number budget, SMA’s models can predict how likely that budget is to be overrun. “If you want greater certainty, you’ll have to have a larger budget.”

Accordingly, the project team decided to choose a budget at the 85th percentile of certainty, which, when the construction contract was awarded, was within 2% of the winning bid. Thanks to @RISK’s integration of both schedule and cost concerns, the Edmonton LRT project began on solid financial footing.

Join decision-makers around the world who

RELY ON PALISADE.