Case Study 20, 23: From Space Launch System, GAO-15-596, July 16, 2015

National Aeronautics and Space Administration’s (NASA) Space Launch System (SLS) program is NASA’s first heavy-lift launch vehicle for human space exploration in over 40 years. For development efforts related to the first flight of SLS, NASA established its cost and schedule commitments at $9.7 billion and November 2018, respectively. The program, however, had continued to pursue more aggressive internal goals for cost and schedule. GAO was asked to assess a broad range of issues related to the SLS program, including examining the extent to which the SLS’s cost and schedule estimates for its first test flight were reliable.

The SLS program substantially complied with five of six relevant best practices, but could not be deemed fully reliable because they only partially met the sixth best practice—credibility. While an independent NASA office reviewed the estimate developed by the program and as a result the program made some adjustments, officials did not commission the development of a separate independent estimate to compare to the program estimate to identify areas of discrepancy or difference. In addition, the program did not cross-check its estimate using an alternative methodology. The purpose of developing a separate independent estimate and cross-checking the estimate is to test the program’s estimate for reasonableness and, ultimately, to validate the estimate. The continued accuracy of the estimates was also questionable because officials had no plans to update the original estimates created in 2013. GAO’s cost estimating best practices call for estimates to be continually updated through the life of the program to provide decision-makers with current information to assess status. Moreover, as stressed in prior GAO reports, SLS cost estimates only covered one SLS flight in 2018 whereas best practices call for estimating costs through the expected life of the program.

Limited cost and schedule reserves placed the program at increased risk of exceeding its cost and schedule commitments. Although the SLS program was committed to a November 2018 launch readiness date, it had been pursuing an internal goal for launch readiness of December 2017, with the time between December 2017 and November 2018 being designated as schedule reserve. The SLS program expected to use a significant amount of schedule reserve, in part to address some technical challenges, and planned to shift its internal goal from December 2017 to tentatively July 2018. This shift would reduce the amount of available schedule reserve from 11 months to just 4 months. In addition, the program planned for cost reserves of less than 4 percent each year and had already allocated those funds for the year, which left no reserve funding available to address unanticipated issues.

Earned value management (EVM) data for SLS remained incomplete and provided limited insight into progress toward the program’s external committed cost and schedule baselines because it tracked progress relative to the program’s internal goals—which had proven unrealistic. EVM data is intended to provide an accurate assessment of program progress and alert managers of impending schedule delays and cost overruns. GAO analysis of available SLS contractor EVM data indicated that the contractors might incur cost overruns ranging from about $367 million to about $1.4 billion, which was significantly higher than what the contractors were reporting—$89 million. SLS was implementing a program-level EVM system that, once complete, would include all contractor work and work conducted in-house by NASA to provide more comprehensive information on program progress relative to internal goals. Tracking to internal goals, however, provided limited information relative to progress toward external commitments. The SLS program lacked comprehensive program-level reporting to alert managers of impending delays and cost overruns to external commitments.

GAO recommended that to ensure that the SLS cost and schedule estimates better conform with best practices and are useful to support management decisions, the NASA Administrator should direct SLS officials to update the SLS cost and schedule estimates, at least annually, to reflect actual costs and schedule and record any reasons for variances before preparing their budget requests for the ensuing fiscal year. To the extent practicable, these updates should also incorporate additional best practices including thoroughly documenting how data were adjusted for use in the update and cross-checking results to ensure they are credible. In July 2018, NASA provided GAO the results of its latest assessment of the SLS’s cost and schedule estimates against its Agency Baseline Commitment. NASA explained how data were adjusted for the updated estimate and the reasons for variances between the original estimate and the current estimate.

GAO reported these findings on July 16, 2015 in Space Launch System: Management Tools Should Better Track to Cost and Schedule Commitments to Adequately Monitor Increasing Risk, GAO-15-596.