Space Fission Power Post #6:Entry-Level Option: Fission Surface Power (FSP) for Mars/Moon

Fission Surface Power (FSP) systems are well suited to be the workhorse of human exploration infrastructure on the Moon, Mars, or other potential destinations (e.g. Titan, large asteroid).  Some potential surface power electrical loads include habitats, in-situ resource utilization plants, rechargeable rovers, construction equipment, and science experiments.  The power output of a single workhorse surface system might be in the range of 20 to 50 kWe, with a lifetime of ~5 to 10 years. In addition, the relatively low-power and “high” mass allowance of a surface reactor might make it the easiest space fission system to develop.  For a human surface exploration mission, fission power has so many advantages over other power alternatives that the mass requirements will not be as stringent.  This allows “mundane” reactor technologies to be used (i.e. stainless-steel, UO₂) at very benign power/flux levels, which makes development simple.  Plus, as part of a “heavy” mission architecture, there should be less programmatic pressure to meet mass targets, or worse, decrease mass during development—a contributor to the downfall of the SP-100 program. 

There are many possible reactor and power conversion technologies available for an FSP system.  Stirling power conversion is generally considered the best option for systems in the 20 to 50 kWe range, with Brayton power conversion systems generally preferred at higher power.  Also, it is a general consensus that a fast reactor using UO₂ fuel and stainless-steel cladding and structure provide the lowest cost and risk system.  Heat transport for this class of Stirling system is via heat pipes or pumped liquid metal.  For a higher-power Brayton system, a gas-cooled reactor should also be strongly considered. These technologies could all be developed affordably; each having specific strengths and weaknesses. The liquid metal systems have issues with pump technology and system reliability, while heat pipe systems have some fabrication and integration issues, and gas-cooled systems are generally heavy.  For more information see FSP_Reactor_Compare and FSP_Cooling_Pros_and_Cons references.

In 2008, NASA and DOE developed a 40 kWe Stirling-based FSP concept that utilized a pumped liquid metal (NaK-78) reactor with a stainless steel and UO₂ fuel system.   This concept was designed with all of the attributes required for an entry-level space fission system.  The power level and lifetime fall comfortably below the knees-in-the curve that could affect development (e.g. irradiation damage, reflector cooling, fuel burnup, passive safety).  For more information on the reactor module see FSP_Reference_Reactor and for the complete system see FSP_Reference_System.  NASA is expected to complete an electrically-heated Technology Demonstration Unit (TDU) in 2013. A successful demonstration will place the system at TRL 6 and ready for flight development. The TDU will also be the first step in a series of non-nuclear tests that, when combined with component irradiation testing, enable a zero-power critical test to be used to qualify the flight system without a nuclear-powered system test.  For more info see the FSP_TDU reference.

Surface power radiation shielding is quite different from a space power mission, where a shadow shield can be used. This is because of scattering from the regolith and spacecraft components, so the shield design usually needs to surround the entire reactor. As a result, surface reactor shields can be very heavy, and can dominate the system mass. Fortunately there are many potential options for using in-situ resources as shielding materials, which can provide substantial mass savings. Regolith could be used in several ways: berming, sandbagging, digging/burying, using natural topography (e.g. craters), or any combination of these. In-situ resources could also be used to fill permanent structures or cans as a fixed part of the reactor structure (e.g. sand could be scooped or vacuumed), and in the most optimistic scenarios in-situ water could fill shield tanks or perhaps concrete could be made. All shielding options that require in-situ resources will depend on significant robotic ability (or existing robotic/human infrastructure) to successfully complete the mission. This could add significantly to mission risk, but the mass savings potential is so large that in-situ options warrant serious consideration. Any option that can pre-deploy and verify in-situ shielding prior to human mission departure would have a significant advantage. A comparison of deployment and shielding architectures can be found in the FSP_Deployment_Architectures and FSP_Shielding_Options references.

The reference FSP concept was designed for use on the Moon, but with the idea of also being used on Mars. There are a few differences between the applications, which are discussed in the FSP_Moon_to_Mars reference. A comparison of the estimated mass of solar and fission lunar/Mars power systems is shown below.

 

Power System	Specific Power (W/kg)
40 kWe Solar + Storage on Lunar Surface (equator)	0.6 W/kg (with 100% power at night) 1.3 W/kg (with 50% power at night)
40 kWe Solar + Storage on Mars Surface (clean surface and atmosphere)	5 W/kg (with 100% power at night) 8 W/kg (with 50% power at night)
40 kWe Solar + Storage on Mars (dust storm, effective insolation 100 W/m2??)	1.8 W/kg (with 100% power at night) 2.8 W/kg (with 50% power at night)
40 kWe FSP system: SS/UO2, NaK-cooled, Stirling conversion.	8 W/kg (with shaped or regolith assisted shield)

These solar power system specific masses are based on 25% efficient, sun-tracking photovoltaic arrays with 50% round-trip efficient regenerative fuel cells.  Despite being closer to the sun, the Moon is in many ways more challenging for solar power than Mars because of the cold 14 days of darkness. The fission system is dramatically better than solar on the lunar equator, and also substantially better at the poles (not shown in the table above).  On Mars, the solar system that provides only 50% power at night is about the same mass, but this does not include the risk of major, prolonged dust storms.  Also, in both cases when higher power systems are utilized, the specific power of the fission system becomes much more attractive.  Ultimately, second generation technologies can provide fission systems with substantially higher specific power.

In summary, the FSP’s combination of mission need and simple development makes it an ideal entry-level space fission system.  As such, this system has seen the majority of work over recent decades. Please refer to the documents referenced within for more detail on the system and the development approach.

Space Fission Power Post #7:Entry-Level Option: Low Power Space Reactor (LPSR) Systems

A space fission system that utilizes a low power space reactor (LPSR) meets the two primary criteria of an entry-level space fission system: 1) a near-term space exploration need and 2) an affordable development approach. The need for a LPSR system is driven by NASA’s need to continue and expand its robotic exploration of the solar system. NASA has dependably relied on radioisotope systems to power science missions for many decades; however, there is no current supply of Pu-238 to meet NASA’s near-term exploration goals.  Because of this, NASA prompted DOE to restart production of Pu-238, although the present effort will produce only a small fraction of NASA’s anticipated needs, and not for several years.  Even if an option existed that could produce all the Pu-238 that NASA would like, the integrated cost of producing the power sources would inhibit the number/level of missions that NASA could pursue. To mitigate the Pu-238 supply/cost issue, NASA has invested in Stirling conversion technology, which can provide ~4 times the electrical output per gram of Pu-238 as compared to thermoelectrics.  This approach appears to be on a successful track, but does not solve the underlying problem.  Even with Stirling engines, a long-term supply of Pu-238 to meet NASA’s exploration needs will require substantial investment, and the program would be threatened continuously by shifting political and programmatic forces.  Furthermore, all of this effort would simply maintain current exploration capabilities, whereas NASA’s historical charter has been to expand our ability to explore space.

As an alternative, a simple fission system based on existing technology could provide robust, long-lived power at 0.5 kWe, 1 kWe, or whatever might be optimal for a specific exploration goal.  The mass of an LPSR system would be similar to a radioisotope system, but more importantly it would likely be of lower cost.  The fission system would be fueled with enriched U-235, which currently exists in substantial quantity and if needed can be made relatively inexpensively. In addition to solving the Pu-238 cost issue, the fission system would also provide NASA with a technology that could grow into a much more ambitious space program; e.g. orders of magnitude more science/communications, outer planet orbiters, Mars surface power, nuclear propulsion, etc.  Fission systems have other advantages, in that they do not have to be cooled during all pre-deployment operations (fabrication, assembly, transport, storage, on the launch pad, etc.), they can load follow power demand, and the lack of radioactivity prior to deployment eliminates the need to mitigate the risk of transport/launch dispersal accidents.

Since the 1970s, there have been several failed US space reactor programs, primarily due to the high cost, risk, and time required for nuclear development and testing. The thermal power of a ~1 kWe system is so low that it can introduce three major advantages.  First, existing technology can be used for all components (and the irradiation environment will be benign).  Second, the burnup of fuel is so small that the loss of reactivity is negligible even over a 10+ year mission, which means that no reactor control is required after initial startup (this improves reliability and decreases development cost).  Even without reactor control thermal power will passively follow the demands of the power conversion system.  Third, the thermal power is low enough to allow a unique testing approach that is not afforded by higher power systems – a low-cost nuclear-powered system test.   This last advantage has the potential to reduce the cost of demonstrating nuclear-powered operation of a space fission power system from ~$1B to <<$100M.  Zero-power critical testing is considered an essential element of all envisioned space reactor programs, as it provides valuable nuclear data at relatively low cost and risk.  The costs of these tests are generally 2 orders of magnitude lower than those estimated nuclear-powered tests of typical space power reactors. At the low thermal power of a LPFS, nuclear-powered system operation (e.g. system startup to full power) might fall within the scope of existing critical test facilities and procedures.

A design approach was developed for LPSR concepts based on anticipated requirements and uses, which is referenced in LPSR_Design_Approach. Science mission power reactors will likely be coupled to either thermoelectric or Stirling power conversion systems. The choice of power conversion might be driven by the “usual” parameters – mass, heritage, reliability, etc., but in this case there might be a strong desire to use Stirling engines in order to keep the thermal power within the scope of critical testing. A concept study of a 1-kWe thermoelectric LPFS was conducted in 2009 by NASA/DOE. The system uses a solid core of U-MO fuel that conducts fission power to Na heat pipes.  The heat pipes travel through the shield to thermoelectric generators. The identical reactor core could also be attached to Stirling engines to produce 3 to 5 kWe. This concept is referenced in LPSR_Reference_System and LPSR_Reference_reactor.

An even lower power reactor concept that provides 500 We via Stirling engines has been developed because a <2 kWt reactor is almost certain to fall within the test envelope of existing facilities. This system directly meets the power needs for near-term NASA space exploration missions, and may be the simplest and most affordable. This concept, referred to as the very low power fission system (VLPFS) is discussed in the LPSR_VLPSR reference.

The only factor that could complicate the development of a LPFS would be a requirement to reduce mass below what could be obtained with existing technology.  In addition to keeping the power system mass low, there are likely to be given stringent dose requirements, which will drive the design to low-mass shielding materials, tight shadow-shielded geometries, and large payload separation distances.  The table below shows the mass of two Pu-238 systems along with 3 possible first generation LPSR systems; the mass in parenthesis includes a shield that reduces dose to allow the use of off-the-shelf electronics (25 kRad).

Engineered System	Alpha (kg/kWe)
120 We Pu-238 thermoelectric (MMRTG)	410
140 We Pu-238 Stirling (ASRG)	180
500 We 1st generation LPSR-Stirling	tbd
1 kWe 1st generation LPSR-Thermoelectric	~420 (770 with COTS electronics shield)
10 kWe 1st generation LPSR-Stirling	~100 (170 with COTS electronics shield)

In conclusion, if the thermal power level can be kept low, and the mass requirement allows existing technology to be used, then a LPSR system offers the lowest cost, shortest path to a nuclear-powered system demonstration (realistically <3 years for <$100M).  A LPSR system also has the nearest term potential usage by NASA, with missions identified that could use the system as soon as it is available.

Space Fission Power Post #8:Second Generation Space Fission Power (SFP) Systems

The previous set blog posts were focused on the need for space fission power (SFP) systems and the entry-level systems that could establish their use in space.  There are numerous potential applications of SFP systems that could be performed by second generation systems; i.e. systems that are hard to develop affordably from scratch, but rather easy to develop after the technology, infrastructure, and programmatic precedent has been established for an entry-level system. One of the main anticipated uses of second generation systems are high power science mission systems that produce 10 kWe to 200 kWe at specific powers of 25 to 100 W/kg. This class of reactor might be referred to as “traditional” space reactors, which has been the focus of two significant US space reactor efforts: SP-100 and JIMO.  These SFP concepts were deemed too expensive to continue development generally because of material issues associated with high temperatures and the need for a nuclear-powered ground test.

First and second generation SFP systems have many potential uses, some of which are included below.

• Ambitious space science and exploration.

-        Mars/lunar/asteroid surface power/propulsion (robotic and manned).

-        Outer-planet missions: Jupiter/Saturn moon orbiters/landers, Pluto orbiter, etc.

-        Interstellar precursor or near-Sun missions.

• Enhanced national and planetary defense.

-        High power and enhanced mobility for defense applications.

-        Potential use for comet/asteroid defense.

-        Synergy with advanced terrestrial and airborne defense systems.

• Significant commercial value.

-        Satellite power, mobility, maintenance, retrieval.

-        Space junk sweeper.

-        Space tourism (orbital, lunar, ?).

-        Eventually, resource extraction and delivery from Moon or asteroids.

The list of candidate reactor and power conversion technologies for second generation SFP systems is very broad, and attempting to summarize these options in a short blog post is impractical. In addition, the best technologies for specific missions are dictated by requirements, and it is hard to discuss/compare second generation technology options in the absence of a defined application.  Finally, it is best to keep our focus on entry-level systems until we establish the use of fission power in space; then we can turn our attention to utilizing the vast potential of space fission power.

Space Fission Power Post #9:Nuclear Thermal Propulsion (NTP)

Solid-core nuclear thermal rockets (NTRs) have the potential for specific impulse (ISP) a factor of 2 better than the best chemical propulsion can offer (NTRs can potentially offer up to 1000 seconds ISP).  This added efficiency can potentially be used to decrease trip times, carry larger payloads, and/or decrease the number of heavy launches from Earth. A significantly shorter trip time is not practical for near-term Mars missions, so the benefit is generally accepted as additional (~2x) payload delivered per kg launched into Earth orbit.  Therefore, the near-term benefit of an NTR is directly correlated with the cost per kg of Earth-to-orbit launches; the higher the cost of launching from Earth, the greater the attractiveness of NTP, and vice-versa.

When compared to Nuclear Electric Propulsion (NEP), the performance advantage of an NTR is higher thrust-to-weight, although with a much lower ISP. The technical advantage of an NTR over an NEP system is that the balance of plant is much simpler for an NTR.  This feature also presents a major disadvantage, because all of the technical risk and development is within the reactor (i.e. nuclear in nature), as opposed to non-nuclear.  Thus, NTR systems cannot utilize the development approach of using a simple reactor (with low nuclear technical risk), and transferring the development risk to non-nuclear technologies and system demonstrations. 

In a solid core NTR, propellant/coolant flows through the reactor core, and thrust is created by expanding the propellant through a nozzle. Hydrogen is usually used as the propellant because the low atomic mass causes the greatest exhaust velocity (i.e. ISP); however, hydrogen is not a preferred reactor coolant because it is highly corrosive, dangerous to handle, and it requires cryogenic storage.  The greatest challenge of an NTR is the extremely high temperature core to provide hydrogen exit temperatures of >2500 K. Materials must be found that have sufficient mechanical properties to support thermal and vibrational stresses at these temperatures while also withstanding possible erosion/corrosion with fast flowing hydrogen. There are numerous fuel forms that can be considered for NTRs, the most extensively studies and tested are the graphite-based fuels of the ROVER/NERVA programs.  In more recent decades, there has been a good deal of research in cermet-based cores and pebble-bed cores.

The component technologies developed during the ROVER/NERVA programs were very impressive and fairly mature, and the existence of the previous work certainly makes the development of NTRs much more feasible than if the program had never taken place, but this does not mean that a modern day development program will be easy.  The fact is that back in the NERVA program they were still testing and developing; and they were in some ways backtracking from system testing to special effects testing of fuel – because fuel erosion was still a major issue.  There was no flight-system ready to fly (certainly not one that would be approved to fly in today’s environment); they were close, but the ability to get back to that point, and proceed from there would be a difficult task.  There are 3 things missing from the old NERVA program that eliminate the potential for affordably developing a NERVA-based NTR. 

1) Physical Infrastructure.  The manufacturing and testing infrastructure that existed at the end of the NERVA program makes any nuclear development capability that exists within DOE or NASA today look miniscule. 

2) Knowledge and Brainpower.  There were hundreds of design, manufacturing, fabrication, testing and system experts working on NERVA in the ‘70s.  Not only are most of these people gone, there has been relatively little knowledge transfer since this was before the era of digital record keeping, and use of the technology has not been continuous.   Back then, nuclear development was more art and elbow-grease, as compared to science and CPUs; they built, tested and tweaked things and used their ingenuity to make things work.  There are many examples of people trying to reproduce products (materials, parts, systems) developed 50 years ago that fail, likely because there were some steps in the process that workers did not record, or simply took for granted (but they would not be intuitive to someone starting from scratch today).

3) Political Environment.  This issue is more than who is in power in Washington, D.C., at any particular moment.  Unfortunately, an anti-nuclear bias is firmly entrenched in the US and the world.  The bias is generally based on an uninformed fear of the dangers of radiation, which unfortunately is perpetuated by the media because exploiting this fear gets ratings (e.g. Fukushima, where there were no deaths and likely no future deaths caused by radiation, but the nuclear disaster got more press and caused more panic than the killing of ~20,000 people by the quake/tsunami).  These fears have made the cost and risk of getting approval to run a nuclear-powered reactor test at least an order of magnitude higher than back in the 1970s.  A program that relies on a nuclear-powered demonstration might have a chance in today’s environment (i.e., a nuclear test that you are confident will work), but one that requires iterative nuclear tests is generally unthinkable without a huge long term cost commitment (on the order of billions of dollars per year for a decade).  In addition to development, launch approval is another factor where politics, public perception can kill a program.  Nuclear material safeguards are a significant issue for all space reactors (except for a small class of low-performance reactors that could reduce the safeguards category).  The difficulty in creating an acceptable nuclear and launch accident safety case is a function of system power and the drive for low mass, and an NTR is much more difficult than a modest power reactor in this respect. Another final programmatic factor that makes NTR development extremely difficult is time - a realistic program will take more than a decade (at least 2x the development time of a simple power system).  Given the realities that occur in Washington D.C., a program that takes over 5 years is risky and a program that takes over 10 years is essentially a non-starter (even the incredibly ambitious Apollo program went from conception to landing a man on the moon is 9 years).

If all 3 of the above issues could be reset back to where they were in the early 1970s then an iterative nuclear test program could be feasible within a realistic NASA program.  However, even if #1 and #2 were still available, #3 makes the costs and programmatic risks unacceptable to any program other than a national-mandate program like Apollo.  The fact is that an adequate human exploration program can be performed with chemical propulsion, so it is unlikely that a nationally mandated NTR program would occur in the foreseeable future.

In conclusion, a nuclear thermal rocket would have significant benefit to NASA (albeit not as much as other fission systems), but it does not warrant more than a small fraction of near-term funding until an entry-level fission system is established in space.

Space Fission Power Post #10:Ultimate Goal: High-Power Nuclear Electric Propulsion (NEP)

A high performance NEP reactor offers more potential benefit to long-term human space exploration than any other space reactor application. The reactor for a high-performance NEP systems would power a multi-megawatt (MMW) system (>>1 MWe) with a very low specific mass (<< 10 kg/kWe). Whereas NTRs offer a factor of 2 performance benefit over chemical rocket technology, a higher performance NEP system can offer orders of magnitude better performance (ISP) – a truly enabling and paradigm shifting technology.  Advances are continuously being made in all of the non-nuclear technologies needed for an NEP system, most importantly power conversion, heat rejection, and thrusters.  Significant advances are being made in ion thrusters (HiPEP) and plasma thrusters (e.g. VASIMR), such that the deployment of high power EP systems appear possible within a decade or two.  Conversely, a reactor for this kind of NEP system is essentially no closer to deployment than it was 50 years ago; which is another testament that non-nuclear system development is vastly easier than nuclear system development.  The difficulty in this reactor is not achieving high power, but delivering power with a very low mass system, which implies very high temperature operation.

A high-performance MMW reactor has some of the same issues as an NTR, most importantly the fuel.  An MMW system will need a very high temperature fuel with a high uranium density (to allow low mass).  Additionally, this fuel will have to go to high nuclear burnup, because every additional U-235 atom burned increases the energy density achieved by the system.  A high temperature structural material that can withstand significant irradiation damage must be found, which is also compatible with the fuel form.  Once these two major reactor issues are resolved, a high temperature, high efficiency power conversion system must be developed that can integrate well with the reactor.  On the plus side, the balance of the system can be developed in a non-nuclear fashion, to some extent utilizing the lessons learned and infrastructure from the earlier development of entry-level and second generation systems.  Even then, the high power density of the core might make resistance-heated system testing impractical, so that a nuclear-powered ground test will be needed; unless several smaller reactors had been previously qualified and operated, and the data from those systems could be used to qualify the MMW system.

There are many potential technologies that have been proposed for low-mass, high temperature MMW space reactors.  Refractory-metal gas-cooled reactors have been seriously considered in previous programs, with high temperature fuel in a particle/pellet configuration.  This reactor could then be coupled to a refractory Brayton power conversion system.  High-temperature liquid-metal cooled reactors (e.g. Li) have also been considered; coupled to potential liquid-metal Rankine systems or perhaps a refractory Brayton system.  Both these options might offer hot-end temperatures between 1300 K and 1600 K, which is high enough to enable a high efficiency thermodynamic cycle while utilizing a high temperature radiator.  For an MMW reactor, radiator size is one of the most important issues, to keep mass low and minimize issues such as deployment, shielding (scatter), and maneuverability. Thermionic systems are a possible candidate, in that they offer very high temperatures with reasonable efficiencies. Molten salt reactors have many attractive attributes (potential for low-mass, long-lived reactors) but a salt would need to be found that can reliably operate at temperatures high enough to make an attractive system. Finally, on paper, MHD power conversion systems could provide very attractive ultra-high temperature performance. Ultimately, this space fission system class (MMW NEP) has several sub-classes within it, and hopefully would evolve to higher and higher performance systems as they evolve.

It is debatable which propulsion system would be easier and more affordable to develop: NTP or MMW NEP. Starting from where we are today, an NTP system might be easier to develop; although both systems are well beyond the bounds of what could be an affordable system. If an evolutionary approach is undertaken for space fission power reactors, then eventually the MMW NEP system would be much easier to develop. Solar electric propulsion (SEP) systems could also serve as precursor NEP systems.  What is not debatable is that the MMW NEP system increases our ability to explore and expand into space much more than an NTP system.