GAO: JWST’s New Launch Window “Likely Unachievable,” Could Bust Budget Cap

The Government Accountability Office (GAO) issued its latest report on the James Webb Space Telescope (JWST) today and the news is worrisome.  Even though prime contractor Northrop Grumman now has teams working on the telescope 24 hours a day, GAO concluded that the latest launch schedule, announced just months ago, is “likely unachievable.”  If the launch date slips past June 2019, the project’s costs are likely to exceed the $8 billion cost cap imposed by Congress.

This is GAO’s sixth annual review of JWST, which is required by the 2012 appropriations act that initially established the cost cap after significant cost growth earlier in the decade. The $8 billion cap on formulation and development has been renewed in each subsequent appropriations act for NASA.  The life-cycle cost of the project is estimated at $8.835 billion, with the remaining $835 million for operations.

JWST is often referred to as the follow-on to the Hubble Space Telescope, though it will study the universe in infrared rather than visible wavelengths.  That will enable JWST to “look much closer to the beginning of time and to hunt for the unobserved formation of the first galaxies, as well as to look inside dust clouds where stars and planetary systems are forming today,” according to its website.

It will not be in Earth orbit like Hubble either.  It will be placed at the Sun-Earth L2 Lagrange point 1.5 million kilometers (1 million miles) from Earth.  It has a very complex design and deployment sequence and cannot be serviced by astronauts like Hubble, so if anything goes seriously awry there is no remedy.  (After Hubble was launched in 1990, scientists discovered a flaw in its mirror that made images blurry.  Astronauts installed a corrective lens on the first of five servicing missions in 1993.  Hubble is still operating today with new instruments and upgraded equipment installed on subsequent servicing missions).

JWST has had a long history of cost growth and schedule delays.  The original  cost estimate was $1 billion, but grew from there.  In 2010, Sen. Barbara Mikulski, then chair of the Senate appropriations subcommittee that funds NASA,  directed NASA to conduct an independent review.  Led by John Casani , it concluded that project management rather than technical issues were the primary problem.

NASA restructured the project in 2011 and developed a new cost estimate that included robust cost and schedule reserves to respond to unanticipated problems that might be encountered.  Reserves provide flexibility to deal with “unknown unknowns” that occur in most technology development programs. A well-managed program will end up with zero reserves at the end.  The cost estimate was $8 billion, with launch in October 2018.

In September 2017, NASA concluded that integration problems at Northrop Grumman would delay the launch to sometime in a 3-month launch window from March -June 2019.  That new schedule included an additional 4 months of schedule reserve.

What GAO found in today’s report is that JWST has used up all 14 months of schedule reserve it had in 2012 plus 2.5 of the 4 months added in September 2017 when the launch was delayed to 2019.  Over the course of the project, GAO said, about 2.5 months of schedule reserve is required each year, but now there is only 1.5 months left with launch more than a year away.  At the same time, the project is in the integration and test phase where problems most often are found.

Based on prior experience, this “amount of reserve will not be adequate for the challenging work ahead, and further delays to launch readiness are likely.”

GAO also notes that Northrop Grumman has not decreased its workforce on the project as planned.  “For the past 44 months, Northrop Grumman’s actual workforce has exceeded its projections and the company is not expected to reduce its workforce until the spring of 2019, when NASA plans to ship the completed observatory to the launch site.”

That means more cost to NASA since this is a cost-reimbursement (or “cost plus”) contract, not fixed price.  NASA renegotiated the contract in September 2017 and agreed to pay $179.9 million more.  That is covered by reserves, so does not exceed the cap, but the concern is about what lies ahead.

For the sixth consecutive year, the JWST project managed spending
within its allocated budget in fiscal year 2017. However, JWST is still
resolving technical challenges and planned work continues to take longer
to complete. Prudent management of its resources allowed the project to
carry into fiscal year 2018 about a third more carry over funding than it
had projected at the beginning of the fiscal year. Program officials said
that assuming the remaining integration and tests proceed as planned
and no long delays are encountered, the existing program resources
accommodate the new launch window of March to June 2019. The project
continues to identify funding options in the event of a delay of beyond the
end of the launch window. Under the 2011 replan, Congress placed an $8
billion cap on formulation and development costs, but any long delays
beyond the new launch window—which, as noted above, are likely—
place the project at risk of exceeding this cap.

GAO did not make any recommendations in this report. It noted that it made recommendations in previous years and NASA followed some of them, but not others — “some of which may have provided insight to the current schedule delays.”

Traditionally, GAO provides drafts of its reports to whatever agency it is reviewing so the agency can present its viewpoint.  GAO publishes the text of the response as an appendix to the report.  In this case, GAO said that while NASA provided technical comments that were incorporated as appropriate, NASA “officials determined that no formal comments were necessary.”