- Status Report
- August 7, 2022
Prepared Statement by Jerry Grey 8 May 2003 (part 2)
Rationale for Continuing Shuttle Operations
There is one overriding reason for NASA to maintain the ability to conduct Shuttle fleet operations even if the OSP is initially successful. Space transportation will remain a high-risk activity for the foreseeable future, so reliance on a single system for ISS servicing (the OSP/EELV) could once more precipitate a crisis much like the present one should the OSP/EELV system fail or otherwise be grounded for an extended period. In simple terms, maintaining a viable second source of access to the ISS ensures its continued operation in the event of a launch system failure. The Russian and European access capabilities could conceivably help to ameliorate this need, but neither can be counted upon, and even if NASA were to resurrect the Alternate Access to Station program, its designs could be a useful supplement, but not the primary ISS delivery system.
Other subsidiary reasons for maintaining Shuttle operational capability are:
(1) It may become necessary to replace one or more of the major ISS elements (e.g., a solar-panel wing), which cannot be carried by any conceivable OSP design;
(2) The sensitive economic situation in Russia (and also conditions in Europe) may deteriorate even further, so that reliance on Soyuz, Progress, and ATV for auxiliary ISS support may become impractical or impossible; and
(3) The Shuttle can provide services and facilities to the ISS that would not be available from an OSP; e.g., extra crew members for major repairs or replacement operations and to help conduct science experiments, water from Shuttle fuel cells, auxiliary equipment for short-term use on ISS research experiments, greater cargo capacity both up and down, etc.
NASA has also pointed out that with an operational OSP the Shuttle could focus on cargo missions to ISS, especially an automated version (discussed later), and could serve as a heavy lifter for future space exploration missions.
The only real negative, of course, is a big one: the additional cost of maintaining the Shuttle fleet in operational status. With a successful OSP available, the Shuttle could be pared down to perhaps one or two flights per year, and possibly even be maintained on a standby basis, flying only when its special capabilities are needed. However, not only would that raise safety concerns, but it doesn’t reduce the required Shuttle infrastructure, which absorbs the bulk of Shuttle manpower and costs.
The safety issue could be somewhat ameliorated by having the Shuttle SLEP program explore reducing the number of crew members and providing the Shuttle with a suitable flight-deck escape capsule, which has been estimated to double the probability of crew survival. The best way to address both cost and safety issues of maintaining Shuttle capability, however, would be to equip the orbiters for fully autonomous operation, including automated docking at the ISS, as the Progress modules now do, and autonomous landings, as the old Soviet Buran did. For those missions in which a crew is needed at the ISS, they could be carried as passengers, as is planned for the OSP.
However, the real justification for continuing Shuttle operations is that the optimum implementation plan for the OSP would be an evolutionary one, as I will discuss later. Hence the Shuttle would be needed at least until the phased implementation of the OSP has been completed. The annual cost impact of Shuttle plus OSP for the next decade under such a plan needs to be established, of course, but the prospect of automating the Shuttle could conceivably reduce that impact, along with annual OSP evolutionary development budgets that are likely to be lower than the annual cost of implementing a fully capable OSP by2012, if the present high OSP development cost estimates are to be believed.
Costs subsequent to 2012 are wholly dependent on the operating cost of the OSP/EELV architecture, which has yet to be even estimated, plus the cost of maintaining the Shuttles in flight-ready condition. Again, the operating cost benefits of a fully autonomous Shuttle should be factored into any trade study of parallel vs serial OSP development, as should all viable alternatives such as dependence on Russian, European, and commercial transport capabilities for both crew and cargo. But until NASA has some idea of the OSP/EELV operating cost, it does not make sense to commit to a full OSP developmental effort aimed at complete Shuttle replacement as soon as the OSP becomes operational.
Summary. In short, both the Shuttle and the OSP (or an equivalent Shuttle substitute) are required for assured access to, and egress from, the ISS. Second-level design requirements for the OSP could focus on either a common vehicle for both crew and cargo or, more likely, different versions having a common technology base. The Shuttle SLEP should include autonomous operation of the Shuttle for cargo functions and possibly also for ferrying crews to and from the ISS.
(4) Given that the OSP program has not yet progressed beyond establishing the Level I requirements, do you think NASA’s plan for spending approximately $750 million on technology demonstrations between FY03 and FY06 is justified? What technologies are the most critical to demonstrate before proceeding to full-scale development?
The primary risk-reduction measure in mission assurance is elimination of single-point failure modes, which is best accomplished by a combination of heritage technologies, proven integrated system health-management techniques, and redundancy, substantiated by test or demonstration and other means of independent verification. A flight demonstration is by far the most effective mission assurance tool. Hence the planned X-37 demonstration program would be highly valuable to OSP development, provided it does indeed address the critical technologies NASA has identified. These include, among others, the thermal protection system; an autonomous, fast-response flight control system; an integrated health-management system, preferably embedded in an fault-tolerant vehicle architecture; and a crew rescue system. The proposed Demonstration of Autonomous Rendezvous Technology (DART) and pad-abort demonstrations are also of high value to OSP development.
It will not be easy to establish which of these technologies are mandatory, to what level of development they need to be brought, the level of development risk, and whether they are consistent with cost goals and OSP operational objectives.
Note that it is not necessary for NASA to wait until the X-37 technology demonstration program is complete before initiating OSP development, especially if the phased development approach I have suggested is used. [Indeed, the NASA plan calls for full-scale development of the OSP to begin in 2004, long before the scheduled completion of X-37 orbital testing]. However, in contrast to Shuttle development, the OSP development program should be structured so that useful technologies and processes demonstrated by the X-37 and the other planned demonstration programs can be readily inserted; i.e., the program should be “drop-in friendly” for new technologies. Again, this is best accomplished via a phased OSP development program.
(5) What design alternatives should NASA examine as it performs its concept studies for the OSP? What changes to the OSP program would you recommend to reduce the cost or accelerate the schedule?
NASA has already suggested that the design trade space for the OSP is essentially open; that is, it could be one or more reusable winged vehicles with passive or active thermal protection and powered or unpowered landing capability, or one or more expendable capsules employing ablative heat shields much like the Apollo capsule, or anything in between. Specific design options must await the formulation of second-level requirements; e.g., mass and dimensions of payload facilities; propulsion and power requirements; the nature of required medical care equipment and supplies; life-support requirements; integrated vehicle health-management system needs; ground facilities; crew escape system requirements; ISS docking, interface, and separation requirements; etc.
NASA level-1 requirements specify that the OSP system must accommodate both rescue and transportation capability for no less than four crew members, although different versions of the system design might be used to perform these two functions. The rescue function must be available no later than 2010; the transport function no later than 2012. NASA’s current proposal suggests that development of both functions be implemented in parallel, at an (admittedly premature) estimated cost of $9 – $13 billion. In the interest of reducing that cost, or at least stretching it out over a longer period to minimize the annual budget impact, it would seem to make sense to develop the required OSP functions serially rather than in parallel.
The urgent need is for ISS crew rescue (which is actually needed by 2006 rather than the specified 2010, in view of the end of Russia’s commitment to provide Soyuz lifeboats for ISS). Why not seek the lowest-cost design approach to meet that requirement and then use the technologies demonstrated and experience gained during that development to develop the transportation capability? There are at least two viable low-cost design options for crew rescue: the original CRV concept and an expendable (or partly reusable) capsule with an ablative heat shield. Other options for use of modified experimental vehicles are discussed below.
Although it might turn out that the transportation capability might indeed require different design features than the rescue capability, NASA should at the very least conduct trade studies on the parallel and serial design approaches before committing to full-scale development.
This evolutionary development option, as well as NASA’s proposed plan, requires an operational Shuttle fleet until the OSP transport function is demonstrated, so the trade study comparing the two approaches should include the Shuttle SLEP options I mentioned earlier, such as fully autonomous operation and a crew escape system. Also implicit in this trade study would be the viability of some means for persuading Russia to extend its Soyuz lifeboat commitment beyond 2006.
In conducting this (and other) trade studies NASA faces the challenge of “requirements creep;” that is, allowing requirements for technical demonstration of the transport phase to affect low-cost rescue options. NASA needs to re-establish cost credibility, and a properly phased, evolutionary program has the potential to do that.
Other Trade Studies. Other trade studies that should be conducted before proceeding to full-scale OSP development include the following:
Basing a Shuttle at the ISS. A temporary cost-saving option that should be explored is to extend the on-orbit lifetime of some of the Shuttle fleet so as to allow an orbiter to remain at the ISS for extended periods, thereby serving both functions required of the OSP. This approach has obvious disadvantages; i.e., it only postpones the requirement for a Shuttle replacement or supplement such as the OSP; it reduces Shuttle operational availability by keeping a third of the remaining fleet inactive for long periods; and it exacerbates the disruption that would occur following another Shuttle loss.
However, it would remove the time pressure on OSP development, especially the 2006 deadline for ISS crew rescue capability, and the presence of the Shuttle crew along with that of the ISS would provide full crew capability for both ISS maintenance and science research; e.g., 10 crew members (or 7, if the SLEP program recommends reducing the Shuttle crew to 4 so as to facilitate crew escape). Also, NASA could reconsider its decision to cancel the low-budget Alternate Access to Station program, whose designs could be evaluated for their ability to supplement ISS cargo transport requirements in lieu of more frequent Shuttle deliveries, especially after Russia ceases Progress flights.
Replacing Shuttle Columbia. Another temporary cost-saving option that should be evaluated is simply replacing Columbia. A four-orbiter fleet, especially if augmented by the Shuttle SLEP, would significantly ameliorate the disadvantages of basing a Shuttle at the ISS. Even if the ISS-based Shuttle option is not pursued, a four-orbiter fleet could allow development of the OSP transportation function to be stretched, relieving the time pressure (and annual budget impact) somewhat. However, without an ISS-based Shuttle the four-orbiter fleet would not resolve the crew rescue function or enable a full crew to occupy the ISS when a Shuttle is not docked to the station. Hence the crew return function for the OSP would still be needed by 2006.
Use of Modified Experimental Vehicles. Modifications that would be needed by the X-37 technology demonstrator or the Air Force’s Orbital Maneuvering Vehicle (OMV) should be costed and evaluated for potential risks as interim solutions to each of the two OSP functions, including the use of multiple vehicles to accommodate the 4-person minimum requirement. Should either provide significant cost reductions vs the OSP without introducing unacceptable risk, this option could reduce the pressure on near-term OSP development. Note, however, that the cost of incorporating a crew compartment could turn out to be prohibitive, even for multiple vehicles.
Other experimental vehicles that could be evaluated for the cost and risk of performing part or all of the OSP function would be NASA’s HL-20 and X-38 or the Air Force’s X-24C. The Air Force has also contemplated developing a generic transatmospheric vehicle, which could be considered as a potential means for augmenting OSP functions.
Evaluation of Apollo-type Systems for both Crew Return and ISS Transport. A top-level assessment of this approach, completed in March 2003, suggests that it might be the lowest-cost option to meet OSP requirements in the shortest time, especially if development of return and transport capabilities were to be conducted serially, as I have suggested. The initial assessment report states, “The (assessment) team concluded unanimously that an Apollo-derived CRV (crew return vehicle) concept appears to have the potential of meeting most of the OSP CRV Level-1 requirements. An Apollo-derived CTV (crew transport vehicle) would also appear to be able to meet most of the OSP Level-1 CTV requirements with the addition of a service module.” This option clearly needs to be evaluated in further detail.
(6) How does the decision to proceed with a design that is totally reusable, partially reusable, or expendable drive design complexity, development schedule, cost, and safety?
Reusability almost certainly implies increased design complexity, a longer development schedule, and increased development cost. The effect of reusability on safety, vis-à-vis expendable systems, has yet to be evaluated. Also, increasing the degree of reusability may or may not reduce operational costs, depending on specific design attributes. It is possible that reusability will, in the long term, prove to be a valuable attribute in terms of operating cost, turnaround time, and reliability, but there is as yet no evidence to support its nearer-term benefit. The often-cited concept of “aircraft-like” operations to realize these benefits requires a full understanding of what is meant by “aircraft-like.” Airplanes are basically designed for cruise conditions while space launch vehicles are designed solely as accelerators. Comparing them without defining the basis of comparison is not realistic.
(7) Can the OSP schedule be accelerated significantly without introducing unwarranted risks? If so, what recommendations do you have?
Once the Shuttle fleet returns to flight status, the urgent need is for crew return capability from the ISS. The evolutionary OSP development program I have suggested would accomplish this goal at the earliest possible time with low risk. The transport capability is not urgently needed as long as the Shuttle fleet is operational, and hence could be developed according to NASA’s proposed schedule, or even stretched out somewhat to reduce both risk and annual budget impact,
(8) What challenges may NASA face in using an Expendable Launch Vehicle (ELV) as the boost vehicle for the OSP? Does the use of an ELV for human spaceflight pose an unacceptable risk?
Safety. The primary challenge is, of course, safety, but that is true for any launch system, not just expendables. The current failure rate (loss of mission) of the partly reusable Shuttle is now 2 in 114, or about 1.75%. The current failure rate of the Delta-2 ELV is 3 in 125, or about 2.4% and of the Atlas 2 – 5 ELV family (including 2A, 2AS, 3A, 3B, and 5) is zero in 64. (That is formally 0%, which is meaningless, but note that the Atlas-5 design failure rate is 0.45% compared with the Atlas-2AS design failure rate of 1.28%, with zero actual failures).
Single-point-failure tolerance is the key factor in launch-vehicle mission assurance. At least one of the EELV systems, the Atlas-5, is claimed to have full single-point-failure tolerance with the exception of its two main engines, the RD-180 first-stage engine and the RL-10 upper-stage engine. However, the RD-180 is probably the most robust large rocket engine ever built (its Russian designers claim it is even reusable), and the RL-10 has proven its robustness over 40 years of operations. Moreover, for components such as engines that are not subject to safe redundancy management, the use of “safe-life” designs and criteria can be implemented, as is common practice for aircraft jet engines (i.e., ground testing to certify design margins with appropriate safety margins). Finally, there are design options for the heavy-lift EELVs which provide engine-out redundancy that would eliminate even these single-point failure modes.
Note that any residual safety risk imposed by using an ELV can (and should) be ameliorated by incorporating an effective crew escape system in the OSP. Such a system (which may turn out to be the whole OSP itself) is likely to be specified in the second-level OSP requirements.
Hence safety is a challenge, but the risk of flying people on an ELV is certainly not unacceptable compared with the partly reusable Shuttle. Also note that the Russian Soyuz launcher, upon which we now rely for all crew-carrying operations to and from the ISS, is expendable, as were the Atlas, Titan, and Saturn rockets used for the Mercury, Gemini, and Apollo programs without a single launch failure.
Recurring Cost. A second potential challenge in using ELVs is the recurring cost per launch (after all, cost reduction was the prime motivation for developing the Shuttle and for creating the X-33 program and the original ISTP). Current estimated launch cost levels released by the Air Force’s EELV System Project Office range from $80 million for the MLV models to $150 million for the HLV models. Although these costs could certainly increase if any special provisions need to be incorporated for OSP operations (e.g., human-rating, if NASA decides not to rely wholly on the OSP crew-escape system), the EELV cost range remains well within the Shuttle’s cost-per-launch envelope.
Booster availability. A third, although lesser, challenge is booster availability. Having two widely different EELV families rather than a single one is definitely a “plus” in avoiding major downtime problems, although there are some cost implications (fortunately not major ones) associated with ensuring OSP compatibility with both families. It will also be necessary to coordinate launch manifesting of the EELV systems with both military and commercial customer demands, but this has never been a serious problem with prior ELV families.
Flight Control Issues. If the OSP design turns out to be a lifting-body or winged configuration, adequate control authority of the EELV booster during transonic flight could become an issue, especially if NASA’s current plan to launch the X-37 technology demonstrator inside a fairing is pursued. The Titan vehicle that was to be used to launch DynaSoar back in the 1960s required the addition of fins for the necessary control authority and a strengthened structure to accommodate higher bending moments. If the EELVs will require comparable “fixes,” there will be cost and schedule implications, which could be exacerbated if no information is available from an encapsulated X-37 flight demonstration. If the OSP design ends up as a ballistic Apollo-like capsule, there will be cross-range restrictions on the return-to-Earth launch window.
That completes my answers to the questions posed in your invitation. However, I have a recommendation for the scenario that NASA should pursue for optimum servicing of the ISS through the completion of its mission, which is estimated to be 2020 – 2025.
The first task, of course, is to resolve the issues surrounding the failure of Columbia and return the three remaining orbiters to service as soon as possible without prejudicing crew safety.
The Shuttle SLEP effort should be initiated immediately, and should include the following elements, to be implemented as soon as possible without excessive disruption of service to the ISS: (1) Converting the 4-person flight deck to an escape capsule suitable for egress during all flight modes; (2) Providing the orbiters with the option for fully autonomous operation; (3) Providing a method for inspecting and, if necessary, repairing the thermal protection system on orbit; and (4) Equipping two orbiters for orbital stays of at least four months. Depending on the availability of adequate budget resources, a replacement could be built for Columbia. Note that during this period, we will continue to rely, to the same degree as prior to Columbia’s failure, on Russian Soyuz and Progress flights and possibly the European ATV.
As soon as one orbiter is equipped for long-term stays on orbit (which should be prior to 2006), that orbiter should be flown to the ISS and based there for four or more months. Until the OSP crew-return version has been demonstrated, the two orbiters suitably equipped should continue to provide that capability, alternating with each other.
Meanwhile, the NGLT program should be pursued and trade studies followed by evolutionary development of the OSP should be conducted, beginning with the crew return function and subsequently proceeding to the crew transport (and possibly cargo transport) functions. (Pending results of the design trade studies, of course, the lowest-cost, nearest-term option is likely to turn out to be an Apollo-derived design). OSP flights to the ISS should begin as soon as the crew return function has been demonstrated, relieving the Shuttles of the need for on-orbit stays.
When the OSP transport function has been demonstrated, the Shuttles should be placed on a standby basis for autonomous operation, to fly if and when needed for lifting large payloads to the ISS, for crew-carrying and cargo-carrying during any OSP standdown, and also for ambitious NASA science and exploration missions in the Solar System.
END OF STATEMENT