Planners and promoters of the 2014 World Cup in Brazil forecast that the event would leave a lasting legacy on the country. So Brazil spent more than $3 billion renovating and erecting new stadiums to prepare for the tournament. Only a year removed, many of the stadiums are unused or underused—one serves as a parking lot for city buses; another attempts to generate revenue through weddings; and others are for sale at cut-rate prices. Few of the forecasted benefits were realized. Many in Brazil now look at the month-long tournament as a waste of resources, and even Rio de Janeiro mayor Eduardo Paes told ESPN FC that "there is regret that we even staged it."

Those are strong words for a country that loves soccer as much as Brazil. But Brazil's experience symbolizes the risks of large-scale infrastructure projects. From the Big Dig tunnel project in Boston to the Sydney Opera House, large-scale infrastructure projects face non-viability when budgeted costs and benefits differ from forecasts. Often called "megaprojects," these construction projects can cost billions of dollars, setting up significant risks for all stakeholders.

Construction of new downstream oil and gas operations, as well as capacity upgrades, in Texas and along the U.S. Gulf Coast face a number of such forecasting challenges. Refiners are expected to add more than 500 Mbpd of new domestic refining capacity in the next five years; and a large portion of this new capacity is expected to occur in the U.S. Gulf Coast region—which already hosts the five largest U.S. refineries. With significant downstream projects in the planning and the construction phases in this region, accurately forecasting the costs and benefits of these projects has never been more important.

Unfortunately, misinformation in forecasting large projects is "the norm," according to Bent Flyvbjerg, an Oxford University professor and expert in the problems and causes of megaprojects. For those considering downstream infrastructure projects, Flyvbjerg's aptly titled, "What You Should Know About Megaprojects and Why: An Overview" in the April/May 2014 Project Management Journal, delivers an unpleasant analysis of megaprojects: "over budget, over time, over and over again." Flyvbjerg calls this the "iron law" of large-scale infrastructure projects.

Various factors inherent to large-scale projects cause inaccurate forecasting. According to Flyvbjerg, his surveys across a 70-year period show that no such projects, regardless of the continent, state, industry, or private or public entity involved, have been immune from these factors. Owner groups must be on guard. The size and accompanying risk of large infrastructure projects is so high that cost deviations due to inaccurate forecasts are proportionally magnified. Yet because of the forecasted benefits of these projects, non-viable projects often move forward even as they risk destabilizing the owner group or wasting resources better spent elsewhere.

Regrettably, due to the inherently unique nature of each large-scale downstream project, there are few "apples to apples" cost comparisons that can be applied from project to project. Because of the shortage of references, forecasting is imprecise.

For some contractors, the potential gains of winning a high-profile project encourages bids that emphasize benefits and de-emphasize costs. Some contractors hope that once they win a bid, they can make up for cost differences over the course of a project, which can affect quality. Or if they are unable to achieve adjustments, contractors turn to owner groups who are unlikely to discontinue a project with millions (if not billions) of dollars already invested, which then affects quality, timeliness, and bottom lines.

Likewise, because some owner groups may lack cost reference examples for uniquely designed large-scale projects, and forecasted benefits are often similar among bids, owner groups often understandably consider the most economical bids. If owner groups engage a bidder who underestimated costs at the beginning, they may find themselves searching for additional financing midway through the project. This intermediate financing can cause instability, including future debt increases, deferred interest payments, and delayed benefits that made the project so attractive on the front end.

Owner groups may contribute to an underestimation of costs if they get drawn into the benefits of the project. The enthusiasm for building a plant with a share of high forecasted profits is enticing. It could raise the profile of a company to the top of certain energy fields. But such enthusiasm could lead to disaster if it clouds objective planning. This scenario is so prevalent that Daniel Kahneman of Princeton University has given it a name: "optimism bias." In Thinking, Fast and Slow, Kahneman describes how project planners "make decisions based on delusional optimism rather than on a rational weighting of gains, losses, and probabilities. . . . As a result, they pursue initiatives that are unlikely to come in on budget or on time or to deliver the expected returns—or even to be completed."

Project planners and promoters must also be aware of overusing "best case scenario" configurations or assumptions to reach price points that will receive approval, and ultimately, funding. "The problem," Flyvbjerg writes, "is that the dubious and widespread practices of underestimating costs and overestimating benefits used by many megaproject promoters, planners, and managers to promote their pet project create a distorted hall-of mirrors." This problem complicates the analysis of which projects deserve funding and which do not, Flyvbjerg adds.

Further, poor initial forecasting often leads to deficient project execution since it can be an indicator of ineffective or flawed planning. For large-scale infrastructure projects, these initial problems are difficult to correct; the interface of components, units, and disciplines complicates recovery efforts. 

However, the outlook is not all doomsday. While the record of success in terms of accurately forecasting costs and benefits on large-scale infrastructure projects needs improvement, there are measures that can help curtail inaccuracies. Owner groups must establish internal procedures to check optimism bias. These procedures should include referencing potentially comparable projects, even as slight as they may be. The process may take downstream planners beyond projects on the Gulf coast, but the benefits are weighty. Appropriate referencing can help to generate baseline estimates and foster accurate forecasting. Owner groups may also enhance accuracy when they engage outside, independent consultants to provide objective reviews of forecasts and bids.

The contract between the owner group and contractor offers another important opportunity to lessen the incidence of forecasting fallacies on the front end. Writing strategic incentives and penalties into construction contracts provide effective measures to reduce underestimation of costs and overestimation of benefits to win bids. Owner groups and contractors should also seek well-designed dispute resolution mechanisms that foster cooperation and curb productivity declines when issues arise. Like incentives and penalties, well-drafted contractual provisions can facilitate the completion of the project—and, if necessary, establish the guidelines for recovering damages when the project's problems cause disproportionate losses for stakeholders.

Large-scale infrastructure projects do not have to end like the World Cup in Brazil, a country so crazy about soccer that it is almost unfathomable to think that hosting the holy grail of soccer events could lead to a sense of resentment for some. However, when decision makers rely on inaccurate forecasts of costs and benefits on such a large-scale, the risk of a let-down is magnified. Accordingly, downstream operators are well-advised to try to take steps to promote accurate forecasting in order to control the costs and maximize the benefits of large-scale construction projects.

This article originally appeared in the July issue of the Houston Business Journal.