Why We Need an Audit of Silver Line Cost, Funding, and Scheduling, Terry Maynard

Last week, the Reston Citizens Association (RCA) Board of Directors approved unanimously a resolution calling for a full audit of the Silver Line—Phase 1 and Phase 2—by an independent entity overseen by a local panel of experts, officials, and citizens selected by our local Congressmen.  In a follow-up post in Reston Patch, RCA President Colin Mills provides important contextual information for that decision. 

At the core of the RCA Board’s concerns is that the construction of Phase 1 is now projected to cost 94% more and be completed some four and a half years later than first systematically projected in the December 2004.  While the Board believes that there will be few additional cost escalations and schedule delays for Phase 1 absent some totally unexpected set of events, its focus is on preventing the same escalations and delays from happening in Phase 2. 

Figure 1:  Cost and Schedule Estimates for the Silver Line, Phase 1, 2004-2011

While a near doubling of cost estimates (based on year-of-expenditure cost estimates) for a four-year project in seven years is discouraging, the estimative shortcomings causing most of the increases seem to be basic forecasting considerations that ought to have been included and/or more accurate at the outset.  A review of the 2007 US DOT Office of Inspector General baseline report on the Silver Line, Phase 1, provides some brief descriptions of the causes for the various price increases:

• In August 2005, a Federal Transit Administration (FTA) consultant projected a $1.8 billion cost based on a 50% preliminary engineering design study, a better understanding of the planned construction approach, and significant changes in right- of-way, materials, labor, and fuel costs.

• In October 2005, FTA’s project management oversight consultant (PMOC) performed an evaluation of the $1.840 billion cost estimate and concluded that the contingency and inflation factors used in the estimate were insufficient.

• In April 2006, a general engineering consultant (GEC) prepared an independent cost estimate based on the true 100 percent Preliminary Engineering design and incorporating revised contingency and inflation factors as recommended by the PMOC. The GEC set Phase I’s estimated cost at $2.065 billion. DRPT subsequently adopted a Phase I budget of $2.065 billion.

• The PMOC found significant issues with the $2.065 billion estimate in a review performed in September/October 2006. The PMOC concluded, “the escalation factor used was inadequate, the contingency was too low, and the project’s cost and schedule could be seriously impacted due to the many uncertainties identified as a result of the lack of design and/or missing scope in the 100% PE design.”

• Based on the March 2007 negotiation of the design-build contract with the Dulles Transit Partners (DTP), the cost estimate was presented as a range of $2.4-2.7 billion because there is uncertainty as to how much the “agency costs,” which are outside of the DTP contract, will be. Agency costs include project management, rail cars, start-up, and testing.

• The 2011 MWAA budget puts the budgeted cost of Phase 1 at $2.756 billion.

Moreover, the Phase 1 estimates never came close to being accurate until a deal was struck with Dulles Transit Partners (DTP), who presumably calculated with some precision all the costs they expected to incur so as not to be caught short as the project unfolded.  And, as recent news reports indicate, even then it appears that there will be further price increases and schedule delays for Phase 1.  (Interestingly, DTP’s bid offer was $2.95 billion for Phase 1, which was negotiated down.  Now it appears that a $2.95 billion cost is well within the realm of possibility.)

One can’t help but wonder why the various sources of error listed in the DOT OIG report occurred and were so extreme.  Indeed, it would appear that most should have been avoided.  The errors raise questions about the environment in which these projections were generated as well as the estimative techniques themselves. 

• Were pressures placed on those preparing the estimates to minimize construction cost expectations to help assure approval of the project and its funding by federal, state, and local authorities?  And are similar pressures being applied now as preliminary engineering estimates for Phase 2 are being drafted?  How can this be prevented or mitigated? 

• Are the methods used by the engineers who put together the various projections sufficiently accurate to be the basis of decisions to approve and finance major infrastructure construction initiatives?  To what extent were established construction engineering forecasting methods used in making these forecasts?  How well have these estimative techniques performed in accurately forecasting major transportation project construction costs in the past?  Will they be better in forecasting the cost of Phase 2?  Do we need a statement of the some probability and/or range of error metric to help understand how to consider a Phase 2 cost forecast?

The unfortunate history of cost escalations for Phase 1 weighs heavily on cost expectations for Phase 2, including whether the current cost estimate for Phase 2 at $3.14 billion (including $371 million in proposed cost shifts to Fairfax and Loudoun counties) with an above ground Dulles airport station are at all believable.  That projection is based on an April 2011 MWAA assessment of the several alternative Dulles station construction options.  It does not claim to be as complete as a “preliminary engineering” study—other than that the preliminary engineering study for the under-terminal baseline study was ongoing—but rather an analysis of technical feasibility, risks, and costs.  At this time, a “100% preliminary engineering” assessment is due in spring 2012.  It will serve as the basis of Fairfax and Loudoun County Board decisions whether to participate in Phase 2. 

What the cost projections for Phase 2 will be, whether they are close to accurate, who will pay those costs, and when Phase 2 construction might begin and end are all still up in the air.  Given the history of Phase 1, the risk is that the total costs for Phase 2 may ultimately be over $4 billion (only 27% above the current $3.14 billion projection), and that toll road users will be paying well over half of that cost—plus interest.  We need to understand fully what the costs are, who will pay them, and when Phase 2 will be completed before we decide to proceed.  An audit is the best way to gain those needed answers.