New Worlds, New Horizons in Astronomy and Astrophysics1 (NWNH) recommended the Wide-Field Infrared Survey Telescope (WFIRST) as its highest-priority large space project, based on the compelling scientific case coupled with the maturity of the design and technologies and the medium-low technical risk and acceptable cost and schedule risk. The decadal survey committee planned the NWNH program in a cost-constrained environment that has worsened since the report’s publication. NWNH placed a strong emphasis on a balanced program of activities that included significant increased investment in the Explorer and research and analysis (R&A) programs, along with investments in technology development and theory. NWNH addressed this climate where funding available is significantly lower than the planning numbers used in NWNH. The decadal survey committee made the following statement to guide plans under this circumstance:2
In the event that insufficient funds are available to carry out the recommended program, the first priority is to develop, launch, and operate WFIRST, and to implement the Explorer program and core research program recommended augmentations. The second priority is to pursue the New Worlds Technology Development Program (p. 237).
1 National Research Council (NRC), New Worlds, New Horizons in Astronomy and Astrophysics, The National Academies Press, Washington, D.C., 2010.
2 NRC, New Worlds, New Horizons in Astronomy and Astrophysics, 2010.
As a result of federal budget constraints and stress on resources resulting from cost growth in the James Webb Space Telescope, the augmentations to the Explorer and core research programs have only been partially implemented. A balanced research program was strongly recommended by NWNH:3
NASA’s core research programs, from theoretical studies to innovative technology development, are fundamental to mission development and essential for scientific progress. . . . Maintaining these core activities has a high priority for the survey committee, and the budget allocations should not be allowed to decrease to address overruns in the costs of large and medium missions (p. 21).
Finding 2-1: If the funding wedge provided for WFIRST/Astrophysics Focused Telescope Assets (AFTA) is not sufficient to accommodate the mission cost and provide contingency appropriate to the mission risk, it could be very damaging to the program recommended by NWNH.
In the following sections, the committee compares the cost risk for WFIRST/AFTA to that for WFIRST as envisioned by NWNH. The committee had available to it the independent cost and technical evaluation (CATE) performed by Aerospace Corporation for the National Research Council assessing the NWNH WFIRST (Joint Dark Energy Mission-Omega [JDEM]-Omega implementation), the project’s own cost estimate for WFIRST/AFTA without the coronagraph, and a CATE performed by Aerospace contracted by NASA assessing both WFIRST/Interim Design Reference Mission (IDRM) and WFIRST/AFTA without the coronagraph. In quoting mission costs for the IDRM and AFTA, the committee refers to the CATE numbers, because those were performed using a common methodology. Due to the immaturity of the design, no credible cost estimate has been performed for WFIRST/AFTA with the coronagraph, although the project provided a ballpark figure for the instrument, integration, and testing, and the additional 1 year of operations.
During the astronomy and astrophysics decadal survey, the independent CATE team at Aerospace concluded that the cost for the NWNH-recommended version of WFIRST, which was based on the JDEM-Omega implementation, would be $1.6 billion (fiscal year [FY] 2010 dollars) and assessed the technical risk at medium-low. The Aerospace CATE performed for NASA estimated the WFIRST/IDRM cost to be $1.8 billion (FY2012 dollars), and the technical risk to be medium-low. The Aerospace CATE performed for NASA for WFIRST/AFTA without the coronagraph estimated the cost for this implementation at $2.1 billion (FY2012 dollars).
3 NRC, New Worlds, New Horizons in Astronomy and Astrophysics, 2010.
Of greater concern than the $400 million cost differential, which is comparable to typical uncertainties for this early design stage, is the fact that the CATE analysis found the implementation risk associated with WFIRST/AFTA to be higher, with a medium rating, compared to the medium-low rating for either NWNH’s WFIRST or WFIRST/IDRM, reflecting the greater technical implementation uncertainties. This implies that the risk for significant cost growth is greater for WFIRST/AFTA than for WFIRST/IDRM.
A number of the primary risks identified by Aerospace are common to both WFIRST/IDRM and WFIRST/AFTA. First, they both employ large infrared focal plane arrays. In the case of WFIRST/IDRM, Aerospace cited detector yield, integration of the 28 individual sensor chip assemblies on the focal plane, and maintaining calibration while on-orbit as challenging. Although the WFIRST/AFTA concept relies on fewer detector chips, Aerospace maintains the wide-field instrument at the top of the list of technical challenges. Fewer detector chips are required in the AFTA-based design (18 chips in a 6 × 3 array), but system integration is still a challenge for an array of this size. Also significant is the fact that the planned H4RG chips for AFTA are a newer technology with unproven production yield and radiation tolerance. Space vehicle fine attitude control was called out as a key challenge for WFIRST/IDRM, and the identical challenge exists for WFIRST/AFTA.
The higher risk rating (medium for AFTA versus medium-low for the IDRM) is driven by two additional risk factors that apply specifically to the WFIRST/AFTA configuration: (1) the risk of mass-growth and the low margin against existing launch capabilities and (2) challenges in the end-to-end testing of the optics system, which is considerably larger than that of WFIRST/IDRM. The low mass margin, assessed relative to the planned Atlas V 541, is a concern at this early stage. Specific areas with mass-growth threats are the wide-field instrument, secondary structures, harness, and the requirement to design for robotic servicing. Additional instrumentation, such as the coronagraph, will only increase this risk. Although launch capabilities will change before WFIRST is launched, the availability of an affordable heavy-lift vehicle is far from assured. Further, the greatest leverage the project has to address any mass problems arising in the early design phase, when all the instrumentation has not yet been defined—namely, decreasing the aperture—is no longer an option due to the inherited telescope asset. The committee, therefore, agrees with Aerospace’s assessment of a higher cost risk for WFIRST/AFTA.
It is the judgment of this committee that there is greater complexity in the design and uncertainty in the cost for WFIRST/AFTA than assessed in the Aerospace CATE, and it results from the use of inherited hardware designed for another purpose. The use of this hardware may avoid some production costs, and it reduces the lead time normally required for a large primary mirror blank. However, the optic is not on the mission critical path, so the lead-time advantage comes in al-
lowing early integration of some components. The suitability of this hardware for the WFIRST/AFTA mission, however, drives other costs and risks.
The gifted hardware does not comprise a complete fore-optics assembly; much of what is gifted has to be modified; and all of it has to undergo the level of structural, optical, and thermal analysis consistent with any new design for a new mission. The need for rework (e.g., regrinding, repolishing, and recoating of the secondary mirror and repolishing and recoating of the primary mirror), as well as thermal testing the hardware, is acknowledged by the project and has been included in the costing. However, the results of the structural, optical, and thermal analysis are unknown at this time, and so the committee recognizes the following additional risks associated with use of this hardware:
- Elimination of descope option. The project team’s ability to manage risk and control cost is limited by the use of the existing telescope hardware. The large 2.4-m optic with its current prescription drives many aspects of the end-to-end system design and eliminates the primary descope option that a project has—namely, aperture size. This lack of flexibility is a particular concern given the low mass margin. The project team has identified several descope options and noted that they will continue to be developed during preformulation and Phase A studies. However, the constraint imposed by the existing hardware makes cost control and cost avoidance more challenging than it is for the WFIRST/IDRM concept. End-to-end optical and integrated payload testing, developing and manufacturing the ground support equipment, and satellite-level integration and test are areas of higher cost and risk for WFIRST/AFTA that are driven by system size.
- Launch loads. The committee agrees with Aerospace that the launch mass margin is a risk area, both with and without the coronagraph, and a likely outcome is a move up to a larger booster such as the Delta IV-H or, possibly, the Falcon Heavy, although the latter is still in development. The project team’s ability to design the WFIRST satellite to envelope the load profiles presented by these vehicles is limited by the use of existing hardware. If this risk is realized, then the project may be faced with redesign of the existing aft metering structure, forward metering structure, secondary mirror support tubes, and/or smaller structural elements. Although the cost of the more expensive Delta IV-H is included in the Aerospace CATE estimate of $2.1 billion (FY2012 dollars), the risk that the gifted telescope hardware is incompatible with the vehicle loads is not included in the Aerospace analysis.
- Operating temperature and thermal design. A major risk to the performance of an infrared instrument comes from emission from the internal components interfering with the faint signals being observed. Heated mirrors are generally inappropriate for infrared instruments due to the thermal emis-
sion from the components. For WFIRST/AFTA, the near-room-temperature operation dictated by the technical design of the inherited hardware is compensated for by the larger collecting area. The underlying assumption, supported by preliminary estimates, is that WFIRST/AFTA will meet the specifications of WFIRST/IDRM at 2.0 microns, offsetting the higher temperature by having a larger collecting area.
Finding 2-2: The use of inherited hardware designed for another purpose results in design complexity, low thermal and mass margins, and limited descope options that add to the mission risk. These factors will make managing cost growth challenging.
Although sensitivity budgets indicate that the requisite performance can be met, there are still many uncertainties at this early stage, and there is very little design margin to accommodate any issues. For example, measured emissivities may prove higher than those assumed due to contamination or other factors, or there may be problems in isolating the warm regions from colder ones, either optically or mechanically. The low margin stems from the fact that the sensitivity is completely dominated by the telescope mirror temperature, and the thermal emission is a very steep function of temperature. The low end of the qualification range of the optics is 277.6 K, and even if the mirrors are operated slightly lower at 270 K, the background on the detectors is still dominated by the thermal emission of the telescope in the F184 filter. If it becomes necessary to run at the 277.6 K qualified temperature, then the thermal emission in the F184 filter will be 2.1 times higher. At 277.6 K, a blackbody emits 73 times more in the F184 filter than the zodiacal light, and with a 2 percent emissivity for each of the two warm mirrors, the self-glow is 2.9 times the zodiacal light, leading to a factor of 2 degradation in performance.
Further, an extension to 2.4 microns was related to the committee by the science team as an option, and this goal would require lowering the mirror temperature to 229 K to have the same ratio of self-emission to zodiacal light. The committee views it as unlikely that 229 K operation can be achieved without significant added cost. With a 2.0-micron cutoff, thermal design margins are uncomfortably tight at this early stage of development, and meeting this may easily become a cost driver. It is also crucial for WFIRST/AFTA to determine the lowest feasible operating temperature that does not drive an expensive qualification program very early in the preparation for Phase B. In this way, the science capabilities, associated design impacts, and system cost can be evaluated with higher confidence.
Finding 2-3: The mission may have to compromise some science performance to ensure that issues associated with the low thermal margins do not lead to significant cost growth and schedule delay.
Finding 2-4: The risk of cost growth is significantly higher for WFIRST/AFTA without the coronagraph than for WFIRST/IDRM.4
The 2.4-m WFIRST/AFTA telescope provides an opportunity to include a coronagraph with compelling capabilities that the 1.3-m WFIRST/IDRM does not. NASA and the project are, therefore, studying coronagraph designs that could be incorporated into the mission and that could achieve the high (109) contrast that would advance some of the goals of the NWNH technology development program. However, the addition of the coronagraph will also increase the cost and complexity of WFIRST/AFTA. Focused effort has been placed by the project on deriving a conceptual design that does not drive the overall mission requirements (e.g., pointing stability and accuracy), apart from mass, power, and mission life, which are necessarily affected by the addition. This is laudable; however, the cost and cost-risk implications of adding the coronagraph are significant.
The WFIRST/AFTA coronagraph was presented to this committee as an instrument with significant technology development components—a significant part of the rationale for its inclusion on the mission. It is currently in the early stage of development. Many of the technologies are immature, and the design itself is uncertain, with three different approaches being pursued. The cost is also very uncertain, given the immature design and the fact that very limited study has been undertaken to assess the true impact on the mission. The project estimates a rough cost for the instrument, its integration onto the mission, and an additional year of operations at $250 million total. However, the cost of increased operations complexity has not been factored in, and, therefore, $250 million is a lower limit for the cost increase. Without a detailed design or an independent cost estimate or CATE analysis of WFIRST/AFTA with a coronagraph, it is not possible to determine whether the addition of a coronagraph would impact the mission requirements (apart from mass, power, and mission life), in which case the WFIRST/AFTA cost with coronagraph could be significantly higher than presented.
Finding 2-5: The coronagraph design is immature, it involves immature technologies, and there has been limited study of accommodating the instrument on the mission. It is, therefore, not possible to quantitatively assess the cost and risk impact to the WFIRST/AFTA program.
In general, technology demonstration missions accept larger technical risk relative to flagship science missions. Their schedules are also more uncertain given
4 See also Finding 2-2.
the uncertainties inherent in developing low technology readiness level (TRL) hardware.
Finding 2-6: Introducing a technology development program onto a flagship mission creates significant mission risks resulting from the schedule uncertainties inherent in advancing low-TRL hardware to flight readiness.
The committee notes that a significant consideration in the NWNH recommendation of WFIRST as its top priority in the large space mission category was the low cost and technical risk. These points are reiterated in the rationale for WFIRST’s ranking in the report.5 NWNH emphasized that added observing modes to WFIRST and expanding the mission scope must be avoided so that balance in the program is maintained and resources can be devoted to other high-priority investments such as the Explorer program and support for basic R&A. The adoption of the AFTA assets has already elevated the overall cost of the program, but more significantly, it has elevated the risk rating and, therefore, the threat of cost growth. The addition of the coronagraph, which elevates the mission cost to $2.4 billion and further elevates risk, creates serious contradictions to the basis on which WFIRST was recommended by NWNH.
Finding 2-7: WFIRST’s moderate cost, low technical risk, and mature design were important to its ranking as the top priority for a large space mission in NWNH. The inclusion of the coronagraph compromises this rationale.
NWNH also ranked the priority of the large- and mid-scale programs in the (now realized) circumstance that the assumed NASA budget wedge was reduced relative to expectations. The implementation of the WFIRST primary science, the Explorer augmentation, and the R&A program enhancements were all ranked above the exoplanet technology development program.6 Neither the R&A nor the Explorer augmentations have been fully implemented.
Finding 2-8: Without corresponding augmentation to other NASA programs accompanying funding to include the coronagraph on WFIRST, the inclusion of the coronagraph is not consistent with stated priorities in NWNH. In a time of reduced budgets, the first priority in NWNH is “to develop, launch, and operate WFIRST and to implement the Explorer program and core research program recommended augmentations.” Implementing the coronagraph would address some aspects of the exoplanet technology development; the exoplanet technology development program was considered a lower priority by NWNH.
5 See, for example, NRC, New Worlds, New Horizons in Astronomy and Astrophysics, 2010, p. 17.
6 NRC, New Worlds, New Horizons in Astronomy and Astrophysics, 2010, p. 237.
Recommendation 2-1: NASA should move aggressively to mature the coronagraph design and develop a credible cost, schedule, performance, and observing program so that its impact on the WFIRST mission can be determined. Upon completion of this activity, and a cost and technical evaluation of WFIRST/AFTA with the coronagraph, an independent review focused on the coronagraph should be convened to determine whether the impact on WFIRST and on the NASA astrophysics program is acceptable or if the coronagraph should be removed from the mission.
NWNH was concerned about the expansion of scope that might make the mission more costly and affect NASA’s ability to implement other aspects of the program. In particular, NWNH stated the following: 7
The committee considers the general investigator program to be an essential element of the mission, but firmly believes it should not drive the mission hardware design or implementation cost (p. 207).
The committee, therefore, considered whether the addition of the Integral Field Unit (IFU) to WFIRST/AFTA would increase operations complexity relative to the slitless spectroscopy planned for WFIRST/IDRM and also considered in NWNH, particularly for supernova follow-up. The planning tasks associated with the IFU are relatively simple, since the supernova fields are visible 24 hours per day, 365 days per year. The time periods for supernova follow-up are dedicated in advance: There will be no other scheduled observations. Once a supernova is identified, the only decision is exposure time. Otherwise, planning tasks are identical to survey observations, with the exposures added to the next command upload. The project estimates two full-time employees in single-shift operations will be required for the supernova follow-up using the IFU, which will take place over a half year of operations.
Finding 2-9: The addition of the IFU does not add significantly to the mission operations complexity relative to WFIRST/IDRM.
In presentations to this committee, the WFIRST/AFTA team offered the possibility of several scientifically interesting operational modes such as “target of opportunity” follow-up of transient events and drift scan modes for fine astrometry that go beyond the simple “shift-and-stare” modes planned for WFIRST as envisioned by NWNH. The WFIRST/AFTA team did not, however, describe or
7 NRC, New Worlds, New Horizons in Astronomy and Astrophysics, 2010.
justify the impact of the increased complexity on operations costs. It is clear that the expanded investigator program, and in particular the addition of the coronagraph, will levy additional requirements on mission operations. In addition, the requisite level of guest investigator support must also be increased to support the expanded scientific scope. The committee is concerned that the operational costs, including scheduling and data management, as well as science support for the diverse observational program, might increase beyond the $42 million per year currently planned ($20 million per year for science and guest investigator grants and $22 million per year for operations).
Finding 2-10: The increase in operational complexity over the nominal NWNH WFIRST concept, which would be required to accommodate expanded guest investigator observing modes and the coronagraph observations, is an additional risk for mission cost growth.