DEMOCRITOS: a spacecraft for 21st century space exploration

Some time ago, ESF and its partners concluded work on the DEMOCRITOS project. It has been quite an interesting two-year effort. DEMOCRITOS is/was an EC H2020 Space-funded Coordination and Support action, with the aim of exploring the high-level design of demonstrators necessary for the realization of a nuclear electric propulsion-based spacecraft, in the 2030-2040 timeline. But what does that actually mean and why did it happen?

What is NEP, and why should I care?

Nuclear Electric Propulsion (NEP) is a means of space transportation where the power to the spacecraft is provided by a nuclear reactor. The reactor heats up a liquid or gas, which, in turn, powers a turbine, which essentially transforms the heat to electrical power. The spacecraft uses electric thrusters for propulsion.

Electric propulsion (EP) is widespread in modern spacecraft (, and there are many companies that offer electric thrusters all over the world. There are many different EP physics for how the thrusters themselves operate. The main point of EP is that it provides very small thrust (less than 1N normally) for a very long time (for years if needed), due to the very high efficiency of fuel use. Such low thrust is almost useless on Earth, but in space, constant acceleration, even at tiny amounts, will eventually build up to very high velocities indeed.

Nuclear Power is, on the other hand, NOT very widespread in space. There are Radioisotope Thermoelectric Generators (RTGs) included in many missions of course, especially in far-from-the-Sun space missions where sunlight is not available in amounts necessary to produce the necessary power levels via solar panels. There have been many flying nuclear reactors in the past, but none remain operational today.

NEP is the thus the combination of a very efficient means of propulsion (EP), with a very powerful means of power production (nuclear energy). It can as you can imagine, offer multiple advantages over ‘conventional’ means of space transportation (i.e. chemical or solar energy). In a very brief summary, there are four main advantages to NEP spacecraft of sufficient power level (MW level):

  1. It can deliver much more mass, and therefore useful payload, to solar system destinations, in relatively short time (at least an order of magnitude increase of useful mass (payload) than current capabilities).
  2. The presence of a powerful nuclear reactor ensures that there is high power availability, in order to power very power-hungry instruments (think ground penetrating radars around Europa/Titan and very large communication through put for the data link to Earth).
  3. It opens the possibility for space tugs, which would carry large cargo to the Moon or Mars, come back, and haul some more cargo. That can be essential in the case of a human outpost, since a large supply chain will need to be maintained and a NEP spacecraft that has the ability to travel back and forth multiple times will be very useful for carrying large volume/mass of needed materials for the outpost .
  4. Finally, a NEP spacecraft will also be able to act as a guardian of the Earth, in the case of dangerous Near Earth Asteroids. If Bruce Willis is not available, one can envisage sending a NEP spacecraft to the oncoming asteroid and parking it around it. The high mass of the spacecraft (they tend to be rather heavy), can act as a small but persistent gravitational force that will nudge the orbit of the asteroid enough to make it miss the Earth (provided that you can actually detect the asteroid on time, NEP spacecraft can get there fast enough and is heavy enough to change the orbit enough to save us all).

That sounds rather nice, so why don’t we have these powerful spacecraft flying all over the place yet?

The need for the DEMOCRITOS project

Unfortunately, power, as always, comes at a price. The spacecraft will have to be very long (to accommodate the radiators, which need to evacuate all that heat, and to keep the nuclear core away from the payload, which ensures that the radiation doesn’t contaminate everything). It will also be very heavy due to the reactor, shield and radiator mass. Large and heavy make for a very expensive and complex spacecraft to build. That is the main reason no NEP spacecraft has been built yet, and the famous PROMETHEUS project of the US, was eventually cancelled.

So, why did the ESF and other Democritos partners go through this exercise, then?

A brief outline of the DEMOCRITOS project is necessary to provide some context here: The DEMOCRITOS project target spacecraft was a 1MWe nuclear-electric spaceship.  The target power level was chosen to represent the jump in order of magnitude in capabilities for the spacecraft and had a nice ring to it (better than 2.5 MWe at least…).

The DEMOCRITOS project was a design project (i.e. no hardware development). It proposed three different demonstrator designs for NEP:

  1. A ground demonstrator design targeting lower power levels (~100KW) in order to evaluate the subsystems of thermal control, power management propulsion, structures, energy conversion, and electrical thrusters that would be needed in such a spacecraft.  The idea is to see if we can create a test facility that can accommodate the aforementioned subsystems working together at the same time. The nuclear core would be simulated by a conventional gas reactor, with its behaviour modulated via software to simulate a nuclear core behaviour. This is important for understanding transient cases. For example when you stop the reactor, a nuclear core takes some time to cool off and thus keeps providing power. If you don’t manage this power, things might start melting and you are going to have a problem. Similarly, starting up a nuclear reactor takes some time and you need to find some power somewhere to start the fluids running around. That is why you need to test the major subsystems together, to understand, at a practical level, the operational procedures, and thereby avoid surprises later on.
  2. A nuclear core demonstrator high level requirements, in order to understand the options of reactor designs available and the issues surrounding the testing of a nuclear space reactor in the ground. As you can imagine, testing nuclear reactors is not a trivial issue, specific facilities (EXPENSIVE facilities from what I learned) need to be built to accommodate a test reactor, you just can’t move it in to a lab somewhere and test it.
  3. Additionally, DEMOCRITOS undertook a phase-0 study (a term meaning a basic design) of a nuclear spacecraft, using the concurrent design facility of DLR (the German aerospace agency) in Bremen. Concurrent design is a technique used often in space industry (and also for automobiles, architecture etc.), where subsystem specialists (thermal, power, structures etc.), sit together with the system engineer/lead designer but also the client, in a room that allows, via computer connections, all subsystems to be updated in real time and influence the others. This makes swift iterations possible, and allows, to a certain extent, a very basic design to be built up to the level of a spacecraft. Our session took 4 days, and we had the privilege to have with us in the room NASA and JAXA engineers as well. The result was the details of the high-level NEP spacecraft design, the 3D model of which is depicted in figure 1.


Figure 1 - A NEP spacecraft with 1MW power level, as designed during the DEMOCRITOS project, at DLR Bremen. This arrow wing version isometric view configuration (for EUROPA mission). Visible is the 90° angle between the four radiator wings to emit about 3 MW thermal energy from the reactor. A circle of solar panels can be seen in the right side, around the payload module, to provide an alternate source of power. Reactor resides at the tip, on the left, surrounded by shielding material.

How does that fit with H2020?

In Europe we have a very advanced nuclear industry and as well as advanced space industry. DEMOCRITOS tried to bring these two communities together to work on an advanced concept for a spacecraft. Most importantly, DEMOCRITOS was conceived in order to create an international community around the concept of NEP and advocate for the uptake of a nuclear spacecraft as a goal for the world’s space agencies in the future.

So what is the interest of having a project like DEMOCRITOS, from a European perspective, since the European Commission was the funding agency behind it?

International cooperation is important

From a non-technical perspective, it is important keep in mind that no European country has had experience with space reactors, although theoretical studies were conducted in the past. Only the US and Russia have flown reactors in space. Thus, understanding in detail what needs to be done in Europe (from a technological/organizational perspective) to even begin the work on such a project is important. The horizon for NEP realization is in the 2030-2040, if it happens, but these things need to be planned well in advance.

The DEMOCRITOS consortium was thus comprised of the European Science Foundation, the French space agency (CNES), the German aerospace agency (DLR), Airbus Safran Launcher (ASL) and Thales Alenia Space Italia. Representing the nuclear industry, the National Nuclear Laboratory of the UK (NNL) was also a core consortium member. Finally, the Keldysh Research Center (KeRC) of the Russian Federation was part of the consortium as well. This is a very important point to be made as KeRC has (contrary to the European entities) practical experience with space reactors.

The usefulness of DEMOCRITOS was thus to open communication channels with non-EU countries interested in the NEP concept. Apart from the very active involvement of KeRC, we had long exchanges with NASA engineers, and representatives of Japan, Brazil and other countries.

Understanding what needs to be done

From a technical perspective, the exercise served as a ‘where are we’ moment. Theoretical concepts have been created before, but DEMOCRITOS was the first time that actual high-level designs were created. The designs serve as a way to take stock in various technologies available to European industry. And this is important because several technologies are, in fact, not available to European entities and only US/Russian industry has access to them. And thus the question materializes about whether Europe is in a position to develop such technologies. The main areas at stake are high temperature resistant materials, liquid metal fluid systems, and advanced robotics. There is no space to go through the technical details in this post, but I will just report here three lessons that were overarching and perhaps not obvious from the beginning of the project.

The first lesson concerns the radiator: on a spacecraft, the main function of the radiator is to keep the system at operational temperature levels, which means essentially to evacuate the excess heat into space.  In our case, the main exercise was how to get rid of the excess heat of the reactor, which was significant. To get 1MWe of electric power, you need about 3.5 MWe of heat from the reactor. Somehow, you have to get rid of ~2 MWe of heat in to space, a non-trivial exercise, which results in enormous radiator structures (the wings of the spacecraft in the figure). This was of course well known from previous designs. What was not known was whether recent advances in technology will make this a bit easier/cheaper. Unfortunately, the answer was no. Due to the high temperatures needed to optimize the system, there is no easy solution for the radiator fluids without introducing high complexity and resulting in a very large volume of radiator. And it creates a very big problem of how to construct this radiator in space, since such large volume cannot be launched into space in one piece (current rockets are too small. We miss Saturn V…). Some ideas that were proposed:

  • launching the radiator alone, to be assembled in orbit by robots or astronauts
  • some sort of 3D printing of the radiator in space
  • use of a different technology to avoid radiator pipes – the so-called “droplet radiator” (

All these solutions either suffer from low technological maturity or introduce very high complexity. We spent quite some time in thermal studies but there is no easy solution, it seems, and technologies that offer a radical improvement, either reducing complexity or increasing performance, are not something that can be expected in the near future.

The second lesson was that, from a regulations perspective, there is a lot of effort to be spent before you are allowed to use such a spacecraft in space. The space regulations for nuclear power sources provide a number of specific requirements for the design of a space nuclear reactor and give a generic design approach based on best practice in safe design. There is a need for clarification on the limitations on nuclear fuel choice imposed by the 1992 Principles for the relevant use of NPS in space (47-68 UN-COPUOS Resolution «  Principles relevant to the use of nuclear power sources in outer space » In this document, it is stated that ‘nuclear reactors shall use only highly enriched uranium 235 as fuel’. This is quite restraining for the available fuels today. The 1992 UN Principles also specifically state that nuclear power sources for propulsive purposes are not covered in the principles.

What this effectively means is that it will require a dialogue with the UN to confirm if these principles should be applied for a NEP spacecraft, where the reactor provides electric power to the thrusters but also powers the other subsystems. This is a legal matter, but a legal matter to be settled in the UN, which takes a very long time (we are talking decades here according to experts). Thus, the lesson is that if you want to do something in the 2030s, you’d better hire some lawyers now.

The final lesson is that a significant challenge will be to convince the European and worldwide public to support the act of putting a nuclear reactor in space... In order to fly a nuclear power spacecraft, the public not only needs to be informed of the project (as is usually and rightfully the case), but it should also be an engaged and supportive stakeholder. But this is not easy to do for a ‘nuclear’ powered spacecraft. Many European countries can be considered hostile to the idea of nuclear power for any purpose and thus the topic is sure to face strong opposition. Considering the case of the European Union and the difference in attitude of the public towards nuclear energy from country to country, advocating for such a project becomes even more complex. In any case, and building on experience from NASA regarding RTG launches, it seems that a very serious communication campaign needs to be undertaken for any effort (10-20% of the total budget….).

In Conclusion

The undertaking of the development of a large, 1MW class spacecraft powered by a nuclear reactor offers significant benefits for space exploration, but at the same time is costly and difficult, and necessitates a long development time. Some rough costs calculated by NASA based on their experience, indicate a cost of around 7 Billion Euro (in 2016 Euros) to build and launch the first mission of the spacecraft, without including the (as yet unknown) costs of developing several key technologies in Europe.

If such an endeavour was to take place, it is certainly going to be an Apollo or ISS level project. There are significant synergies in technology development with the nuclear, aeronautic and defence industries, with other possibilities for technology transfer between sectors that deal with high performance electronics and advanced materials, so it is possible to ‘share the costs’ of technology development. Explicit coordination actions on these common areas are just beginning.

In my mind, it makes sense IF you assume that exploring the solar system further (with better capabilities) is a worthy cause. There is no technology right now that can provide the amount of power that nuclear power can for space. At some point in the far future, we might have fusion reactors or something else very exotic, but chemical or solar sources will never be able to match the potential of a fission reactor for power.

If you make that assumption, I really think that a NEP spacecraft is the way to go for the 21st century. If you don’t make that assumption, and rather believe that space exploration does not need to advance in big steps and that slow and steady is the way to go, then the cost and effort is perhaps difficult to justify. There is no “right” answer to this question and whether space exploration is a worthy cause is also up for debate (to be addressed in a future post). I strongly believe that it needs to happen though.

Europe should at least explore this option, as well as other options of similar audacity and élan, that can place, for the first time, Europe at the driving seat of a large initiative in space. Europe is a space power, with great successes under its belt but has never led, thus far, a programme big enough that can be seen as really pushing the boundaries of what humanity can do in space. To forge a common identity, you have to do big things as an entity. Opening the path to the solar system might be a good way to go about it.


Header image: Nanedi Valles valley system on Mars - ESA/DLR/FU Berlin (G. Neukum)


Leave a Reply

Want to join the discussion?
Feel free to contribute!

Leave a Reply

Your email address will not be published. Required fields are marked *