Competition and the future of the EELV programby Stewart Money
|
| Though not a product of the original EELV program, the Falcon 9 presents an opportunity to rejuvenate the program with the type of competition that was originally intended but never achieved when the DoD implemented the two-vendor strategy in 1998. |
For the last five years, United Launch Alliance (ULA), the joint venture between Boeing and Lockheed Martin, has been the sole source provider of launch services to Department of Defense (DoD) and National Reconnaissance Office (NRO) for medium and heavy lift. Despite steadily rising launch costs, the company has maintained its position based on the performance and reputation of its two vehicles, Atlas V and Delta IV, and the lack of viable competition. That position is now under threat, principally from SpaceX and its own EELV-class vehicle, the Falcon 9. With two successful flights under its belt, and a third high-stakes flight now scheduled for February, the Falcon 9 offers a compelling alternative to the ULA products both in terms of pricing and its 100% US production.
Though not a product of the original EELV program, the Falcon 9 presents an opportunity to rejuvenate the program with the type of competition that was originally intended but never achieved when the DoD implemented the two-vendor strategy in 1998. It is also an opportunity to judge the merits of a new approach to launch vehicle development, foreseen by many but only now brought to life by SpaceX, versus that of the taxpayer-supported monopoly of ULA, which has produced reliable launch vehicles, but at an unsustainably high price. Hanging in the balance is the future of world’s largest launch purchasing program and the critical affiliated government and commercial markets that it invariably influences.
The immediate point of contention is whether the Air Force will proceed with a proposed “block” purchase from ULA, or will it allow new entrants to bid on medium- and heavy-lift launches conducted for the Air Force and NRO, and if so, how quickly and to what extent. It is a decision framed by the emergence of two distinctly different government reports that cast in stark contrast the differences between SpaceX and ULA.
The first report, released in August, was a presentation of the results of a comparative exercise by NASA using the NASA Air Force Costing Model (NAFCOM) to determine what it might have cost the space agency to develop the Falcon 9 using either a traditional NASA development culture, or one which more closely represented the commercial approach taken by SpaceX. In the initial round, both calculations utilized a cost plus fee format, and measured Design Development Test and Engineering (DDT&E) for the Falcon 9 and one flight vehicle. The results were an estimate of $3.977 billion under a NASA model versus $1.695 billion following a commercial approach.
Noting that SpaceX’s actual costs were nowhere close to one to the model’s output, NASA personnel visited SpaceX headquarters to understand the difference and refine the inputs. The second round, which by this time included numbers for developing the Falcon 9 plus two flight vehicles, resulted in an estimate of $1.3 billion for a NASA development model with a cost plus fee contract, versus $443 million for SpaceX using a fixed fee approach, a number which more closely matched the actual costs cited by SpaceX.
The numbers themselves are not actual budget estimates, and of course are subject to interpretation, but what stands out is that in either circumstance, the SpaceX approach was much cheaper, coming in at only one third that of a traditional development method. What is perhaps more insightful is the explanation provided by the company for its exceptionally controlled costs, and what it says about the rest of the launch industry as a whole. According to the presentation, SpaceX ascribed its costs to three primary factors. The first was the controlled expense of a lean, production-oriented workforce. Essentially, the cost of the Falcon 9 was a product of the total workforce expense of the company, which presently employs about 1,600. The second factor was the reduced organizational complexity of a vertically integrated operation, highlighted by the fact that SpaceX estimates every dollar sent out of the company represented three to five dollars in additional costs. Consequently, SpaceX cites a figure of 80% internal production value for the Falcon 9. The final factor, reduced infrastructure costs associated with the DDT&E effort and a high utilization percentage, closely correlates to the first two.
While it seems painfully obvious that the factors cited in the presentation represent at least one successful strategy for controlling costs in an industrial environment, it is certainly at odds with that taken by the traditional launch vehicle industry. ULA member companies Boeing and Lockheed Martin, as well as Orbital Sciences Corporation and ATK’s proposed Liberty launch vehicle, all depend heavily upon outsourcing for major components.
| With the Falcon Heavy at least a year away from first flight, however, and certainly several years from passing a multi-flight threshold, purchasers and policymakers have to work with what is currently available instead of what might be. |
What is perhaps most interesting about the results determined by the model is that they tend to correspond with the real launch cost disparity between SpaceX and ULA, which is at the core of the EELV debate. In the closest apples-to-apples comparison based on vehicle performance, which is a Falcon 9 versus an Atlas V 401, the SpaceX vehicle is priced for 2013 on the company’s website for all the world to see, at $54–59.5 million for a commercial flight, while a government flight would also include mission assurance fees of around 20%. Pricing for the comparable Atlas is much more obscure, but is frequently cited in a range of $150–180 million. A full accounting, however, would have to include a portion of the billion-dollar-a-year launch capability subsidy that ULA receives under a second contract, which would add at a minimum another $120 million to the price, bringing the total to a staggering $270–300 million. Even under the most conservative estimates, the Falcon offers a three-to-one price differential.
For the moment, at least, ULA’s advantage lies in the fact that the current Falcon 9 does not possess the capability of launching the larger defense payloads. That advantage may prove fleeting, though, as SpaceX both upgrades the Falcon 9 with Merlin 1D engines, and moves closer to the first flight of its Falcon Heavy, which poses a mortal threat to ULA as it offers a launch capability greater than the largest ULA product, the Delta IV Heavy, at a price which is still less than the smallest Atlas V 401 configuration.
With the Falcon Heavy at least a year away from first flight, however, and certainly several years from passing a multi-flight threshold, purchasers and policymakers have to work with what is currently available instead of what might be. Consequently, the SpaceX foray into the EELV market comes against the backdrop of an intergovernmental agency effort to better coordinate launch service acquisitions as a means of grappling with rapidly escalating prices, particularly from ULA. In March 2011, as part of draft Memorandum of Understanding among DoD, NRO, and NASA regarding launch vehicle plans, DoD served notice that it intended to proceed with a previously-reported plan to commit to a large, multi-year purchase of 40 launch cores from ULA, while continuing to work on criteria for eventually allowing new bidders to enter the program at a later date. Given the size of the purchase it was undertaking, as well as the tepid enthusiasm for prioritizing new entrants, it was hardly a ringing endorsement for the values of competitive bidding. NASA, for its part, politely declined to participate, citing a different budgeting process for its decision to pursue launch providers through its own more flexible NLS II contract, which allows an annual “on ramp” for new participants as well as a defined risk-based criteria for entry. With SpaceX one test flight away from making operational trips to ISS, the Falcon 9 is already in the NASA stable, certified as a “medium risk.”
The proposed Air Force purchase was the culmination of an intense lobbying effort on the part of ULA, who, clearly hoping to forestall any competition, had been heavily promoting the massive block buy as a means of running out the clock before any new system could go into effect. The proposal called for the Air Force to purchase five launch cores per year, with NRO responsible for three. With an estimated budget of $15 billion over five years, the program’s basic numbers worked out to an average of $375 million per core. While the plan appeared to escape major attention throughout the spring and summer as the nation focused on the final shuttle flights, and Congress sparred with NASA over the Space Launch System, the onset of fall brought what had to have been unwelcome scrutiny in the form of a Government Accountability Office (GAO) report highly critical of the proposed block buy, and which questioned the USAF’s ability to make an informed decision based on the limited information available.
| Which is the more likely alternative, that SpaceX would develop a successful launch history with the Falcon 9, or that ULA would achieve a 65% price reduction in the price of its vehicles? |
Among the findings, the rationale for the block buy was based on a poorly-substantiated claim by ULA that the launch industrial base would be endangered if was not allowed to proceed. This claim however, relied on a survey ULA submitted to its own vendors, complete with a cover letter suggesting how they should respond in order to support the block buy and “enhance their collective business.” GAO characterized the survey as “neither designed nor administered in a manner consistent with sound survey methodology procedures.” Also noteworthy was part of the questionnaire that was intended to assess short- and long-term risks, in which ULA supplied the responses because, as one company official bluntly stated, “we wanted certain answers.” That’s a tactic that, in the common discourse, is sometimes referred to as cheating. In addition to questioning the validity of the survey, GAO also noted the company was at a loss to provide details as to why its costs were drastically increasing for many key components, including engines. Equally troubling was an inability to explain what actually constituted “mission assurance charges” despite the fact that they can make up a substantial portion of a given launch’s costs, leaving the buyer without any means of determining the relative value of the charge.
In questioning the number of boosters to be ordered under the plan, the GAO report found that ULA was requesting nearly twice the number of cores that DoD had actually been using during the previous five years, and that it would be committed to purchasing five new cores for each successive year, regardless of the actual usage rate. Despite a current backlog of launches due to payload program delays, this approach would in all probability result in storage and refurbishment charges for unused cores. The result, conveniently enough, would be to limit launch opportunities for new entrants (read SpaceX) possibly well beyond the five-year time frame. GAO also warned that with much of the cost increase blamed on uncertainty in the SLS program, DOD was running a risk of locking in unnecessarily high prices even as that program began to take shape.
The GAO report was quickly picked up by Senators John McCain and Carl Levin of the Senate Armed Services Committee, who issued a joint letter on October 21, advising the US Air Force to hold off on its purchase decision, at least for now. Considering Senator McCain’s well-publicized scrutiny of previous large defense contracts, most notably Boeing’s proposed tanker lease deal, this could not have been a welcome development.
For its part, the Air Force appeared to distance itself somewhat from the original block buy proposal and pointed to a October 14, 2011, announcement of a New Entrant Certification Strategy, as evidence of its sincerity in opening up EELV procurement to new entrants. The new strategy would be similar to that developed by NASA, and would seek to develop a risk-based matrix for integrating new entrants while maintaining assurance for its most valuable payloads. On November 7, the USAF released its detailed plans for allowing new bidders, officially titled the United States Air Force New Entrant Certification Guide, and announced plans for an industry day to discuss the details.
These developments, though welcomed by SpaceX and appearing to open the door to competition, did not clearly resolve the issue of the block buy, which is still under consideration. Three days after it received the Senate letter, the Air Force responded by requesting that ULA provide alternate pricing scenarios based on a smaller buy beginning with a minimum of three rather than five years, and with annual purchases ranging between six and ten cores per year. Given the close ties between the Air Force and its major defense contractors, as well as its apparent willingness to proceed with the block buy despite the obvious problems highlighted in the GAO report, it may be difficult to judge to what extent the competition is really being opened up until specific contracts are announced.
One person who remains unconvinced is Senator McCain, who followed his letter to the Air Force with more direct action in the form of two amendments to the National Defense Authorization Act of 2012, both of which are intended to bring the EELV program under closer scrutiny. The first would reclassify the program as development rather procurement, and the second would force it to defend decisions which conflicted with recommendations in the GAO report.
| With NASA, for better or worse, focused on developing the Space Launch System, it falls upon the EELV program to be the primary agent of change in picking up the mantle of reduced cost that it carried once before. |
In a 2004 interview in Aviation Week regarding the potential impact of SpaceX on the market, Michael Griffin—a year before becoming NASA administrator—observed that the legacy aerospace contractors, including his former employer, Orbital Sciences Corporation, needed “existence proof” that the path SpaceX was then undertaking could succeed. What was for years a theoretical discussion over whether a new launch vehicle provider could offer a better alternative than the status quo became something much more real when SpaceX successfully launched the first EELV-class Falcon 9 last June. It took on added emphasis when SpaceX repeated the success with a second flight in December. Now, approaching the eve of an all-important third flight, much of the debate has already been settled.
The one remaining issue—and it is an important one—is whether the Falcon 9 can achieve the reliability necessary to launch the most critical national defense and scientific payloads with the same degree of confidence that is bestowed on the Atlas V and Delta IV. This quite rightly is a feature not to be haggled away over price, particularly for one-of-a-kind type missions such as the just-launched Mars Science Laboratory. Nevertheless, the massive cost differential between SpaceX and ULA, as well as the obvious inefficiencies in the cost-plus-fee format highlighted in both the NAFCOM model and the GAO report, present a conflict that must be resolved in favor of both the future and the taxpayer. It is time to ask the question: which is the more likely alternative, that SpaceX would develop a successful launch history with the Falcon 9, or that ULA would achieve a 65% price reduction in the price of its vehicles, while foregoing its launch assurance subsidy?
The Space Shuttle and EELV programs were both designed to reduce launch costs. The shuttle, pursuing partial reusability, failed completely on that measure. The EELV program, born out of that failure to serve the military’s needs, followed a much more conservative path through building a better expendable. To a degree it succeeded, at least in regards to the more expensive and less reliable expendables it replaced, but progress stopped at that point. From almost the moment the contracts were announced in the late 1990’s, costs have steadily increased and, as the budget for the proposed block buy indicates, are set to increase even further, even if ULA retains its monopoly position.
The effort to maintain that monopoly is based on a line of reasoning which, like any good fiction, requires a willful suspension of disbelief. In this case, ULA’s contention that a block buy is necessary because the industrial base is threatened asks one to accept the proposition that SpaceX, the Merlin engine, and the Falcon 9 booster are somehow not a part of the industrial base, whereas the Atlas V’s Russian-supplied RD-180 engine is.
While ULA officials would be loathe to admit it, the industrial base is clearly stronger now on account of SpaceX than it has been at any time since the joint venture was formed. To the extent that foreign-sourced engines are acceptable, however, the impending debut of Orbital Science Corporation’s Taurus II rocket, also powered with Russian engines, marks the addition of another vehicle with potential application for the EELV program, further undercutting the ULA rationale.
With NASA, for better or worse, focused on developing the Space Launch System, it falls upon the EELV program to be the primary agent of change in picking up the mantle of reduced cost that it carried once before. Fortunately, this time around, there is no heavy lifting required, no rocket to develop, and no launch subsidies to pay. The only requirement is that it manage launch opportunities in a reasonable and equitable manner with a weather eye to the future.
As opposed to perpetuating the ULA monopoly for another five years or longer based on a false premise, and only then opening up a new phase of competition, the single most important step that the Air Force could take to actually ensure the stability of the launch base for the long run is to permanently transition to open and fair bidding on a per-launch basis as quickly as possible, while protecting the most critical payloads. The simplest path would be to follow the precedence that is already being set by SpaceX and require the completion of a limited number of demonstration flights that are not funded through the EELV program, followed by several funded certification flights with test or low-value payloads. Given the cost disparity between SpaceX and ULA, the transfer of even a few flights in the 2015–2017 time frame, apparently already reserved for ULA as a part of the block buy, would more than offset the cost of a satisfactory series of certification flights in the near term. Furthermore, considering the Air Force’s avowed interest in exploring flyback boosters in next decade, it would seem one obvious goal of the test flights could be to gather data on the “rocket back” maneuver SpaceX is already contemplating as part of its Falcon reusability program.
An accelerated transition to competitive bidding would not only reduce program launch costs, but would also eventually increase the total number of new launch opportunities. Most significantly, the creation of an open and predictable purchasing environment for this highly influential market segment would further embolden new participants as well as encourage the introduction of innovative improvements from existing launch providers. It could, in fact, more than offset mounting budget pressures that are limiting future opportunities for both military and civilian space missions, and spark a new era of low cost, high frequency launches for all users. On the other hand, if it finds a way to ignore the opportunity presented by SpaceX and the Falcon 9—and others that may follow—the US government will have officially thrown in the towel and cynically accepted a future which can only be described by the term we have all come to associate with the age… downgraded.