NASA’s Kilopower Program (0:00-14:07)
David Poston’s entrance into a career in space reactor development and how NASA’s Kilopower
project came to be
Q: Did your career start out with an interest in nuclear or in space?
A: David Poston is the Chief Reactor Designer at Los Alamos and the Reactor Design Lead for
KRUSTY. David originally went to Los Alamos to design space reactors and get them in space.
He holds a Bachelor’s and a Master’s degree in mechanical engineering. After graduation,
David worked for General Electric. Once he caught the space nuclear bug, he knew it was what
he wanted to do. The first space reactor project he worked on was the SP-100 government
effort in the 1980’s. After the project’s failure, David decided to get his PhD in nuclear
engineering and to strive to come up with ways to do it simpler and easier, leading to the
building and testing of the first new reactor in 40 years. Lots of interesting reactors were built
and tested in the 1950’s and 1960’s to try novel nuclear ideas. The Nevada National Security
Site, which is the old Nevada Test Site for weapons testing, allows for some experimentation,
but is still under tight regulation. David designed his test to be operated within the safety
authorization basis, working with the regulator to get through the process. Idaho National Lab
has some desert land available, but Nevada has the biggest area that is remote from the public.
David worked on a project for NASA called Kilopower, which is a very compact, lightweight
power source to provide 1-10 kilowatts of power. This would be enough for early human
missions and a lot of science missions. This provided the ability to test at the device assembly
facility on the Nevada National Security Site at the National Criticality Experiments Research
Center (NCERC). The facility takes things critical from a nuclear reactor standpoint to get a
sustained nuclear reaction. At the facility’s authorization basis, David determined they could run
at 5 kilowatts of thermal power for a day inside the facility and learn everything about how the

reactor operated. David is against moderators, only because the testing capabilities are not
there. The reactor is a chunk of highly-enriched uranium metal about the size of a paper towel
roll. The uranium is surrounded by beryllium oxide, which is the best neutron reflector in nature,
allowing the reactor to be very light and small. Beryllium oxide scatters neutrons very effectively
and has a high atom density, providing a very high probability of the neutron being returned.
This reactor can’t go critical or produce radioactivity unless it is surrounded by the beryllium,
increasing the safety of the reactor. The reactor has to be pretty small for the reflector to be
worth it in terms of radioactivity, so it is not usable at commercial scales. David’s test produced
5 kilowatts of thermal power; within the fuel, the thermal power is conducted through the metal
to heat pipes. Heat pipes were invented in Los Alamos for this application, but have since been
adopted for a wide variety of uses. Heat pipes are used to transfer the heat from the core to the
power conversion system, which are Stirling engines. Commercially available Stirling engines
existed, but they were not space qualified. The fission power is generated right into the metal
and conducts into the heat pipe. The heat pipe has liquid sodium inside to transport heat at a
high temperature, approximately 800 degrees Celsius. The heat boils the sodium down at the
core region, the evaporator region, to create vapor pressure. This pressure pushes the heat up
to the Stirling engine where it condenses and gives its energy to the heat pipe wall, conducting
to the Stirling engine. Getting the liquid metal back to the core is tricky without gravity, so a
wicking mechanism is used to draw the material down.
Stirling Engine Reactor Design (14:07-22:45)
The importance of simplicity in nuclear reactor designs and why David Poston committed to the
use of a Stirling engine in his reactor prototypes
Q: Since the core of the reactor is a giant chunk of metal, is the centerline temperature a lot
hotter than the outside temperature?
A: David Poston is the Chief Reactor Designer for the NASA Kilopower KRUSTY reactor
experiment, which has a large chunk of uranium metal in the core. The power is relatively flat
across the reactor because it is so small and there is not as much peaking as seen in larger
reactors. A temperature gradient is created to conduct the heat from the inside to the outside,
with approximately a 20-30 degrees Celsius delta T from the inside of the core to the heat pipes
because of the relatively low power and uranium is a good conductor of heat. The biggest
problem is failing a heat pipe because redundancy is needed. Heat needs to be able to be
conducted to the adjacent heat pipes so heat can always be removed. This reactor operates in
steady-state and self-regulates. The reactor control never had to be moved during the test. Only
0.1 percent of the fuel is burned, allowing it to last decades to centuries at this level. The life
time limiter is the Stirling engine and NASA is doing reliability testing, which comes back to
redundancy. The first mission can either show that they can last 10-15 years or at what rate
they start failing and have to change the design to get more life time. All of the space reactors
that have flown in the past used thermoelectric conversion, so the whole system was in a steady
state, but the problem was efficiency. A Stirling engine can provide 25-30 percent efficiency of
thermal heat conductive electricity, but thermoelectric systems can only provide 4-5 percent
efficiency. David’s conclusion is the reactor needs to be made as simple as possible because
there is not as much ability for testing. KRUSTY (Kilopower Reactor Using Stirling Technology)

was the prototype of the 1 kilowatt Kilopower. The KRUSTY test was almost identical to the
Kilopower reactor, but the difference was how the heat was rejected from the Stirling engine.
This exact technology can be used for 1-10 kilowatts. David did look at the chance of getting
more life time out of the core by switching to low enriched uranium because of fuel swelling
possibilities, which would bring the unit up to 20 or 30 kilowatts. Neutronic simplicity and how
the reactor operates is the most important aspect of the design because it is the hardest thing to
test. Power systems up to 5 megawatts have been designed that use the same physics, but the
fuel has to be changed from uranium metal to uranium oxide and the power conversion cycle
has to change from Stirling to Brayton. Once a system gets about 100 kilowatts, a Brayton
conversion system is much more efficient. KRUSTY provided the first real data from a new
nuclear reactor in 40-50 years and the test was almost exactly as predicted.
Applications of Nuclear Reactors in Space (22:45-34:06)
Future applications of the KRUSTY reactor experiment and why nuclear power is vital to space
colonization
Q: What did the KRUSTY reactor experiment evolve into?
A: The KRUSTY reactor prototype that David Poston developed was a successful test of a
Stirling cycle nuclear reactor. Near term, the Kilopower reactor based on KRUSTY as-is
provides one kilowatt for a NASA deep space mission, with some applications in orbit. One near
term project could be a 20-30 kilowatt lunar lander. Another great application for a 10 kilowatt
reactor is nuclear electric propulsion. Jet Propulsion Labs (JPL) recently published a paper on
the possibilities, which includes orbiting around anything in the solar system. Delta V, or how
much acceleration the vehicle has, is limited by the mass that can be launched on the rocket is
known. The most important parameter is specific impulse, which is comparative to miles per
gallon, looking at how much acceleration change can be achieved. With nuclear electric
propulsion, enough acceleration change can be achieved to slow yourself down to possibly
leave one orbit and go to orbit another moon. Chemical propulsion is not very effective in space
due to a lack of oxygen. The shuttle had hydrogen and oxygen combined, but most of the mass
was oxygen. A majority of this oxygen is spent in the first few minutes of launch. There is five
orders of magnitude more energy density in uranium than there is in hydrogen oxygen. A
nuclear thermal rocket runs a reactor extremely hot, up to 2500 degrees C and hydrogen is
flowed through it. Since it is so hot, the exit velocity provides a much better specific impulse, or
miles per gallon, than chemical propulsion. Using a nuclear reaction for takeoff was explored in
the 1960’s and 70’s, but standard rockets have gotten so good that it wouldn’t make sense to
use nuclear electric propulsion except to launch out of Earth’s orbit. Direct fission propulsion has
a very thin fuel element in which the fission product escapes, where an electromagnetic field
takes the fission product and puts it in the right direction. Right now the space station has 100
kilowatts of power with solar power that is working very well. There are three places in which
solar is not very good, but where nuclear can be. The first is on the Moon. The Moon has 14
days of darkness, making it very cold and storing electricity becomes more difficult than a
nuclear reactor. Mars is further out from the sun, making solar panels less effective, and the
dust storms reduce the solar insulation, or amount of heat flux, so much that the solar farm that
would be needed to power a base would be huge. A Kilopower reactor would be the equivalent

of football fields of solar panels, so the reactor would be much lighter to get to Mars. The third
place where solar panels do not become effective is past the asteroid belt and Jupiter. David
Poston imagines Kilopower scaling up; the Kilopower philosophy and physics could go up to 5
kilowatts of energy. Human colonization of the Moon or Mars might require 20-30 kilowatts per
person. Once the colony is self-sufficient, native materials could be used to build power plants.
The 5 kilowatt reactor barely fits on Elon Musk’s Starship, so until bigger rockets are used, this
would be the largest reactor sent to space. It could also be shipped in pieces once some
infrastructure is in place in the colony.
DUFF and KRUSTY Reactor Experiments (34:06-53:52)
A summary of results from the DUFF and KRUSTY reactor experiments and what made them
successful programs
Q: Tell me more about the opportunities and challenges of working in the lab system.
A: David Poston has been at Los Alamos for 25 years, initially going there to design space
reactors. David proposed the DUFF test in 2012 to show NASA they could still do anything.
DUFF was a 10 watt system that used an existing reactor and ran enough heat pipe to power a
Stirling engine. KRUSTY went very smoothly with the help of Pat McLure, who was very good at
knowing what the regulations were and integrated the regulator, the Department of Energy
(DOE) very quickly and early on. The team set up a program with gates in which they would
show the regulator they know what they are doing, predict the results, and move forward.
Advanced reactor concepts are being developed, but the difficulty the Nuclear Regulatory
Commission (NRC) and DOE have is that only the people that designed them have the tools to
evaluate them, since they are such unique concepts. David’s system with the regulator allowed
him to make predictions, do simple, safe and small steps, and wait for the okay to go onto the
next step if the results were predicted accurately. This worked in place of the traditional system
of getting the regulator’s tools to match the predictions. The whole process, from design,
prototyping, and electrical testing, took three years. The regulator usually had six to seven
people involved in the meetings. David used the MCNP code to calculate transport and the
CINDER code to calculate burnup. This was used to gain a source term and sent to Pat McLure,
who knew more about how aerosols are formed, transported, and dispersed. David’s team had
to calculate and predict how long KRUSTY would be on the criticality machine, Comet, which is
used for lots of important missions. They determined how long it would be for the radioactivity to
be low enough for someone to go in the room, how long it would be until it could be physically
taken apart, and how long it had to sit disassembled before a post-irradiation examination. This
test was solely to measure the magnitude of radiation via neutron fluxes and gamma doses. A
lot of the data during and after the test was gathered from dosimeters on people and was
matched to predictions. David warns those with a burgeoning interest in microreactors to keep
things simple. Dozens of failed programs have tried to take too big of a technology jump. There
must be a realistic amount of time and money to get things done. Oklo might be the right team
doing the right thing; they are using a technology similar to Kilopower. Almost every project out
there now is doing things too complicated. NuScale’s reactors are the most doable, but the
question is whether they will be efficient enough to be economic. David has gotten cynical about
big projects and prefers to take the small steps. He would rather ask for $150 million to put a

reactor on the moon instead of putting money into studying other things. The $150 million would
fund a KRUSTY-like program, not a traditional government reactor program. The small team
was why KRUSTY worked; four people were in the core team, including a mechanical designer,
power system specialist, safety specialist, and reactor specialist. NSA oversaw the project with
NASA and allowed the small team to make small decisions. Space nuclear is important because
curiosity and exploration make being a human fun, interesting, and worth living. Science and
technology also benefit from space programs. The more we explore and discover, the more we
can answer some of the world’s questions and become enlightened. Finally, space nuclear is
important to our viability and sustainability. It is human responsibility to keep the “light of
consciousness” alive and establishing colonies on another planet is an insurance policy for the
human race.