You are not logged in.
Antius,
1. If you launch from the ocean, you can fly a truly enormous rocket. Nova, a 15m diameter Saturn V with 8 F-1 engines, is the maximum size rocket that the cape was designed to accommodate from an infrastructure and community planning standpoint. If the rocket produces more than 11Mlbf, and ITS greatly exceeds that level of thrust, then you'll have to launch from somewhere else or pay for destruction of surrounding private property due to the sound pressure waves emitted from rocket engines that produce more thrust. That is why ITS will never launch from the Cape unless surrounding residential properties are purchased, moved, or upgraded with windows that won't be shattered by the sound pressure waves produced by a launch.
2. A single turbo pump can provide propellants to multiple nozzles, which is exactly how the Russians do it because they never learned how to counteract thrust instability in higher thrust engines, but then the turbo pump becomes a single point of failure. There's no way to "shut off" an engine that malfunctions because the turbo pump is connect to all the engines or so many engines that there is no engine-out capability.
3. What you're proposing would weigh more, cost more, and be less reliable than simply mounting the second stage atop the booster. The Falcon 9 and Falcon Heavy rocket design is likely as simple and reliable as they can be. If SpaceX developed a full-flow staged combustion LOX/RP-1 engine for the upper stage and combined that with composite tanks to offset the higher mass, since Merlin Vac is already a unique production item, that'd probably be the best we can hope for in terms of performance. Naturally, that'd add to the cost of the rocket, but maybe the little boost in performance FFSC provides would be enough to maintain current levels of performance with full reusability.
Advancements in the area of in-space propulsion like fusion drive (Dr. Slough) or EMDrive (Dr. Shawyer) would eliminate any desirable characteristics that spaceships like the one atop ITS could provide. It'd be best from a cost, complexity, and mass standpoint to just deliver everything to low orbit, then have systems with functionally unlimited specific impulse take over from there. The fusion drive is 5100s using Lithium metal and the impulse drive is only electrical input power and ultimate system durability limited. Both systems have few moving components, mostly being based upon applications of electrical power to produce thrust, with plasma jet power being used in the case of the fusion rocket and microwave power being used in the case of the EMDrive.
I see. There are no magic bullets it would seem.
Ocean launch of very large rockets has always seemed like a good idea to me. The old sea dragon idea of building rockets in dry dock and floating them to launch site. Increasing scale is usually a good way of bringing unit costs down, although total cost still increases.
One way of reducing launch vehicle performance requirements within the scope of existing or near term technology would be a skyhook. The upper stage would only need to accelerate the payload to some fraction of orbital velocity. The hook catches the payload and momentum is exchanged. Solar electric propulsion can then be used to raise the orbit of the counter weight back to its original position and the process begins again. This is by its nature a long-term investment, as the tether and counterweight have at least 30 times the mass of the payload being lifted.
If solar electric propulsion can be used as the propulsion system for transfer vehicles between LEO and LMO, then regolith from the Martian moons could be used as reaction mass.
Some questions that have often intrigued me:
(1) Just how big could we build a liquid fuelled rocket? Could we build a 1000te to LEO launcher? How about 5000te? These things must be subject to economy of scale?
(2) Since turbo-pumps are not generally reusable, could we build a multi-engine rocket using a single turbo-pump? That is to say, a single large pump serving multiple combustion chambers?
(3) Space-x have developed a reusable lower stage. Why not contain the upper stage in a payload bay within the lower stage? That removes the need for explosive bolts, thermal protection and the need to design the upper stage to resist atmospheric drag.
The intermittency issue is a fair one to raise. It clearly can be disruptive. That is more of an issue with wind than solar, which is more dependable within in a range. The intermittency issue has not yet been solved but that doesn't mean it won't be. There was a time when solar was hugely more expensive than nuclear. That is no longer the case. I think storage will go the same way. There are many potential technical solutions to the storage issue. Eventually one will take the lead and be fine tuned, till combined with solar it becomes competitive.
There have always been solutions to intermittency; they involve storing the intermittent energy in some way and releasing it in a dispatchable way. The problem is that they are capital intensive and end up eating a large chunk of the energy due to inefficiencies. So you get hit both ways. On Mars, the situation is worse in some ways, as you are starting out with only 43% of the sunlight that you get on Earth. So your solar power system has a 2.3x longer energy payback time. Adding in storage makes things worse.
I don’t think solar is a good solution for Mars, but here is my best stab at what I think will minimise its disadvantages:
I think it is correct to say that intermittency is less of a problem if it occurs in predictable cycles, but I think the answer lies, if there is one, in low tech solutions. If solar dynamic power plants are built on Mars, then heat can be stored in cheap materials like rock and used to provide 24/7 power. That way you at least only have to build one power plant and oversize the collector area to cover 24 hour generation and thermal losses from the store.
The embedded energy in the storage system could be kept relatively small, as materials used can be unprocessed natural materials like rock which need to be corralled around heat transfer pipes, but otherwise not processed in any way. Because of the near vacuum conditions, ordinary regolith should be a good insulator for a thermal store. Other ways to get around poor EROI would be to operate at the highest possible temperatures, use high heat capacity and low vapour pressure coolants (liquid metals) in the collectors and use condensing direct cycles for power generation if you can. The high temperatures push up Carnot efficiency, condensing direct cycles avoid the need for energy invested in heat exchangers, heat exchanger temperature drops and condensing cycles reduce pumping power. So basically dish type solar collectors, probably using a liquid metal coolant, feeding into a thermal store and direct cycle S-CO2 power cycle drawing energy from the store and then generating electric power. Use carbon steels where possible and try to avoid thermal transients within the plant. To maximise the efficiency and EROI of a thermodynamic power plant make it as big as you can. Also, try and keep the whole thing as compact as possible, so run the S-CO2 generating plant at high pressure.
This is a relatively large scale solution, but dynamic cycles with thermal storage would seem to be the best chance of getting round poor EROI. If temperatures in the store can get up to 500°C, the efficiency of the power generation cycle might reach 40%.
In terms of energy on Mars, though, a wide range of different considerations apply.
If we are talking about a small colony of say up to 1000, I don't think there is any prospect of them building their own nuclear reactors. That means you are condemning the settlers to continue importing tonnes of energy equipment from Earth. Having looked into it, I see no reason why with imported 3D printers, purification machinery, glass and other manufacturing capability, that the settlers shouldn't be able to produce their own PV panels on Mars from ISRU materials like silicon. They can easily make their own chemcial batteries as well.
We have already discussed Mars-made PV in another thread. The energy economics are not favourable. These systems are not energetically favourable here on Earth, which is why Germany is having the problems that it is. Things are even worse on a planet with less than half the sunlight. The energy payback time would be several years before you even begin to think about energy storage.
The problem with Mars built Magnox reactors is not their technological sophistication, it is their size. Anyone that can build a solar thermal power plant has the technological sophistication to build a graphite moderated natural uranium reactor. The minimum practical heat output of a Magnox reactor is about 150MW. A Mars colony is unlikely to need something that big until its population is in the thousands.
That said, it is difficult to see why importing a small nuclear reactor from Earth would be burdening the colonists. If the thermal power plant is manufactured on Mars, then the only thing that needs to be imported is the reactor core and control systems. For the SP-100 that is about 1100kg for a unit capable of producing 2.5MW of heat. A larger fast reactor in the tens of MW range would be even more efficient. One kg of fast reactor fuel taken to 10% burn-up will yield 720,000kWh of heat and 288,000kWh of electric power at 40% conversion efficiency.
Regarding nuclear, if you do have a radioactive leak, isn't it way more dangerous in the dusty Mars atmosphere - not being washed away by rain?
Possibly. The source term is the same. Any release plume would be less buoyant in the Martian atmosphere and would cool and descend more rapidly, but Martian winds could spread radioactive dust a long way.
I also think nuclear power is too labour intensive. The PV panel manufacturing process on Mars can be largely automated and once deployed, the panels can more or less be forgotten. Robot vehicles can give automatically blow away dust off the panels.
PV panels are a poor solution, as already discussed. All thermal power plants require some level of manning, whether they are solar or nuclear. The power density of a nuclear power plant is higher, so there is good reason to believe manning levels will be comparable or lower.
Even if we continue to import PV panels, it is much easier to stow those in rockets rather than nuclear power plants.
You really think so?
The problem is that we are comparing apples to oranges. Dispatchable electricity from a coal or nuclear power plant is not the same product as variable, intermittent energy from a wind or solar power plant. In order for intermittent renewable energy to do the same job as dispatchable energy, it must be backed up by a conventional power plant and excess power at peak times must be exported, stored or wasted in a dump load. Without back-up and storage, energy production will not meet demand. If that doesn’t happen, the result is frequency transients that will crash the entire grid.
The renewable energy basically gets dumped onto the grid when it is available. Conventional power plants must then cut back production, losing market share, and then ramp up production when renewable electricity falls off load. The power plant must sit there, fully manned and in hot standby, essentially waiting – whilst unpaid capital, maintenance and labour costs stack up. The only economic benefit that renewable energy sources provide is to save fuel in the back-up power plants. For high renewable penetration you would need storage as well. So we need three power plants instead of just one. Germany provides a cautionary example that we should think twice before attempting to follow:
http://www.telegraph.co.uk/finance/news … onomy.html
http://www.theenergycollective.com/will … ys-economy
A large chunk of renewable energy production in Germany is hydropower, something that has been in place for a very long time. Hydropower is able to compete with fossil and nuclear because it does not suffer from the same disadvantages as intermittent renewables. It is relatively power dense and is highly controllable (dispatchable). It is no surprise then that it is already maxed out in most countries, with little opportunity for further expansion. If we had the ability to ramp up hydropower, no one would bother with intermittent renewable energy sources.
Establishing new coal and nuclear power, faces some severe hurdles in developed countries. For one thing, very little of either type of power plant has been constructed in the western world for the past 25 years. Most new power plants have been natural-gas combined cycle gas-turbine plants. The expertise and workforce needed to construct large steam plants no longer exists in most western countries. Deindustrialisation has robbed us of the capability. To get the sort of low cost electricity from nuclear power as is seen in countries like France, large scale economies are needed in the production of identical units. On top of that, any nuclear project in the US or UK is basically strangled by a regulatory regime that stretches out build times, which ramps up capital costs.
Comparing a large-scale PV solar against new nuclear in levelised cost studies, will therefore tend to make solar look a lot cheaper than it really is (i.e. back-up and storage costs are rarely included) and nuclear a lot more expensive than it really needs to be (real costs will depend upon the capability of the workforce, scale effects and the regulatory regime).
Of course, there is no coal or natural gas on Mars, at least not that we know of and no oxygen to burn them even if they do exist. The wind carries 100 times less energy. The sun is 2.3 times weaker. Water is frozen as hard as stone, so prospects for hydropower are not promising. Biomass means growing things in greenhouses that you need to burn instead of eating. If these things don’t work well on Earth, how well do you think they will work on Mars?
Nuclear power on the other hand is likely to be more efficient on Mars. Heat sink temperatures are lower, boosting Carnot efficiency and I doubt that there will be any such thing as ‘waste heat’ on Mars. The only question in my mind is whether a Martian economy should import its reactor or build them from local resources. I raised the prospect of Magnox because it is something that can be built at a relatively low level of industrialisation - no enrichment is needed and components are relatively simple and can be made from low grade materials such as plain carbon steel, graphite and magnesium-aluminium alloys. On the downside, these reactors are inherently large, low power density and have reduced thermal efficiency compared to other more advanced technologies. But if someone were to ask me to build a nuclear reactor with limited materials and manufacturing capability, this is the way I would go. If I had enrichment capabilities, heavy water, zirconium, stainless steels and reprocessing plants, the answer would be different.
Why not use a LOX-rubber hybrid? Solid propellants sound downright dangerous. The sort of hobby that could end up depriving a man of his arms.
According to this site, the Candu heavy water reactor core has a power density of 11MWth/m3.
This compares to about 0.5MW/m3 for Magnox. So a heavy water reactor core would definitely be smaller for the same amount of power. This is largely due to the much better moderating power of heavy water compared to carbon.
In terms of thermal efficiency, they are about the same. A Candu is more difficult from a materials viewpoint, as we need heavy water for the moderator, zircaloy for the pressure tubes and fuel cladding and stainless steel for primary circuit pipework and steam generators.
For a Magnox, we need carbon steel for the boilers, diagrid, and the stressing tendons, a lot of pure graphite for moderator, concrete for the pressure vessel and magnesium alloy for the cladding. Magnox is technically easier, but the core is much larger.
Even though Mars is colder than Earth, I gather experts in settlement construction are concerned about eliminating excess heat. A greenhouse will be receiving nearly half a kilowatt of sunlight per square meter all day; that's a lot of heat. So I don't think greenhouses will need to be heated, nor will living areas.
For a settlement buried underground that would certainly be true. Also, ordinary Martian regolith at 6mbar pressure, is about as insulative as rock wool here on Earth.
Greenhouses at night are a different matter. A greenhouse at a temperature of 288K (15C) will radiate heat into the Martian sky at a rate of 390W/m2. It is doing that 24 hours per day. During the day, it will gain an average of about 300W/m2 for the 12 hours that the sun is about the horizon. So, it will require heating about 75% of the time if it is to remain warm around the clock. The heat load may be reduced by using a reflective cover at night and at times when the sun is low on the horizon, but it is difficult to see how the greenhouse could remain above freezing without some heating.
Hello Louis, I mean the later. In my calculations I have assumed that any power system deployed on Mars will need to produce energy around the clock. So if a solar panel is producing XkWh per day, then time averaged power is (X/24) kW.
Louis is correct in this case. The EROI of nuclear energy in most parts the western world may be lower than calculations would suggest, because of the labour required to implement safety management at all stages of a nuclear project. People and their lifestyle energy consumption needs to be part of an EROI calculation if those people are working on what is being examined.
Nuclear energy can be very cheap or very expensive. It all depends upon the adequacy of the design, the build time of the reactors and the regulatory regime. It is always easy to push up the cost of a new project, by adding additional requirements.
Euan Mearns wrote an excellent article for Oil Price magazine on the cost trends for new nuclear reactors and the causes of recent price increases.
http://oilprice.com/Alternative-Energy/ … Power.html
Also, the article he references appears to be open access, for those that are interested:
Here is a detailed description of the Magnox type reactor.
http://www.iaea.org/inis/collection/NCL … 052480.pdf
The French developed a similar reactor - the Natural Uranium Gas Graphite (NUGG) reactor.
Whether it would be easier to enrich deuterium on Mars and develop a pressure tube reactor, is something worth discussing. I do know that heavy water has a slightly higher boiling point than normal water. It may be possible to use nuclear waste heat to enrich it using a cascade of boilers and condensers. Maybe this could be done at a small scale.
Well into the base building phase on Mars, the most practicable option for power supply on a Mars base will be nuclear reactors and solar power systems imported from Earth. However, as base population increases beyond 1000 and the scope of local ISRU increases, the increasing mass requirements of the power supply will become increasingly burdensome. A Mars colony will need large quantities of power to mine resources, reduce metal ores, manufacture machinery and other goods, new living space and propellant. Heating loads on Mars will be at least equal to those of colder regions on Earth and greenhouses must be heated to prevent frost damage to food crops at night.
Whilst solar power systems will provide energy in niche applications, the limited EROI and low energy density of these systems will make them relatively expensive and living on Mars will be energy intensive, as most products, food and living space must be manufactured locally. For this reason, high living standards and high growth rates can be sustained only by using nuclear reactors.
The options for nuclear power development at colony phase are either to (1). Import high-enrichment, high power density reactor cores from Earth and build secondary systems on Mars; (2) Attempt to enrich uranium on Mars and build high power density cores; (3) Build natural uranium reactors on Mars.
The first involves a large import bill, although the energy density of enriched uranium is extremely high. At 10% burn-up, it comes to 30GW-days per tonne.
The second would involve centrifuge enrichment of uranium hexaflouride, which would appear impractical for a small colony. On Earth, the smallest nations to attempt this have populations of several million people.
The third would involve natural uranium reactors, using natural uranium mined from ores on Mars. This would appear to be the only intermediate term option to direct import of nuclear fuel. Candu reactors fission natural uranium using a deuterium moderator. However, deuterium enrichment would appear almost as capital intensive as uranium enrichment. The other option is graphite moderated reactors, with a low cross-section cladding (Magnox). These have the advantage that graphite is easily manufactured on Mars and the reactors are quite easy to build. The UK began construction of its first magnox in 1953 and completed it in just 3 years. However, power density is low compared to light water reactors.
https://en.m.wikipedia.org/wiki/Magnox
The minimum critical diameter for a magnox reactor was calculated as being 26 feet, but this would generate negligible power. Calder hall was the first UK power reactor, with a diameter of 36' and a power output of 50MWe (182MWth). The core contains 120 tonnes of natural U and 1140 tonnes of graphite.
US per capita electricity consumption is 12MWh per capita per year. Assuming a Mars colony consumes energy at the same rate, a magnox reactor with power output 50MWe would provide enough power for a colony of 36,000 people. The waste heat from the reactor would be sufficient to heat 450,000m2 of greenhouses to a temperature of 300K around the clock. On this basis, we might begin building such a reactor when base population reaches several thousand.
The answer will basically determine which Mars mission architecture is practical at this time.
If, as would seem likely, the world's space agency's fail to get their arses into gear and develop a space nuclear reactor, future Mars missions will need to depend on Solar.
If whole-system power density comes in at >15W/kg for a baseload supply (i.e. including storage) then future missions could proceed with Mars Direct style missions using solar power. If whole system mass is closer to 5W/kg, then it makes more sense to build a solar power system at a single base and have all new missions go to that base. Long range expeditions to distant parts of the planet would be dispatched using long range rovers.
For ultraflex / megaflex, specific power is listed as 150W/kg at 1AU, according to this JPL link.
http://www.lpi.usra.edu/opag/meetings/a … uchamp.pdf.
At Mars, sunlight intensity is 0.43x what it is at 1AU. If atmospheric transmittance is 0.7, and time averaged powr is (root-2)/4 times full sun intensity. That gives a time-average power of 16W/kg.
I am assuming that the goal is to provide 24/7 power, as this is needed for life support and is highly desirable for all functions. This requires that somewhere between 67-75% of power consumed is stored, as the sun is only above the horizon 50% of the time and solar irradiation will follow a sine function of intensity during daytime hours. Lets be generous and assume only 67% of power needs to be stored. If storage efficiency is 33%, then 14,675kg of cells are needed to supply a constant power of 100KWe.
The energy density of hydrogen-oxygen storage system is taken to be 250Wh/kg. Some 1800KWh must be stored. So storage system mass is 7200kg. Total system mass is therefore 21,875kg. That's a mass power density of 4.57W/kg. That is a little less than one quarter of the mass power density of the SP-100 with stirling power conversion. In this configuration, SP-100 is a less than optimal system as it must operate at part power. Perhaps Louis would like to check my calculations?
On the subject of SP-100: SP-100 was cancelled largely because there was no foreseeable need for it in the late 90s, as NASA had no realistic humans to Mars or moon programme. The Clinton administration were also keen to exterminate all of the US advanced nuclear projects for political reasons. The integral fast reactor programme was eliminated by the same people. It was close to commercialisation at that point.
If we needed such a reactor, the programme could presumably pick up where it was left off. How much it would cost to develop is a subject of some debate. That said, SP-100 was a thing of the 80s and 90s. We could probably improve upon it today, as there have been some advancements since then.
The 11.4kg/kW figure comes from NASA, but you'll see it all over the place in space power distribution systems design. This has traditionally been a relatively accurate estimate of what most PMAD systems for high-output applications end up weighing. IIRC, the 7kg/kW figure I used previously came from the PMAD for the DS-1 probe. However, that was a much lower output application.
88W/kg is still impressive. On Mars that would still give power output of 26.5W/kg for the panels in full sun and 9.4W/kg on a time averaged basis.
If a 33% efficiency is assumed for LOX-hydrogen electrolysis/ fuel cell energy storage, and 75% of energy is stored, then effective power density of the panels drops to 3.76W/kg. If the energy storage mechanism is 200Whr/kg, then total mass required is 9000kg. The total mass of the panels is 26,596kg. So total system mass is 35,596kg. If the panel mass is halved, them total system mass is 22,298kg, or 4.5W/kg.
By comparison, the SP-100 with stirling cycle would yield 19W/kg. But the SP-100 would suck up a billion dollars of development funds and would face endless regulatory project delays. So maybe that extra 25 tonnes of power supply weight for a solar power system is something we can live with. What matters is total project cost after all.
Antius,
That mass figure you quoted seems incredibly high. No current production thin film technology requires that kind of copper mass for power transmission. I haven't seen any thin film PV technologies that come close to ATK's MegaFlex or DSS's MegaROSA, as it relates to array power density. There may be some sort of graphene-based thin film technology in a lab somewhere, but in the real world MegaFlex and MegaROSA win the power density argument. ATK's UltraFlex has flown and will fly in the near future on real robotic missions to Mars. Incidentally, a number of solar power companies have started integrating the power inverters into their arrays so that less in the way of power conversion equipment is required and losses are reduced.
In the real world, panels would need local transformers as well as inverters to step up voltage. But as I noted, if transmission distances are a few metres or less, it isn't an issue. When systems get to be 100m across, transmission requires increasing voltage. That wouldn't be an issue in most applications. But transformers are heavy.
Here is a link to the tech spec for Alta Devices thin-film PV. This is a California based company that specializes in producing lightweight PV cells for applications like electric aircraft.
http://www.altadevices.com/wp-content/u … -brief.pdf
Take a look at the 5x cell module. This has an open circuit voltage of 5.47 volts. Under normal circumstances, no one would produce a panel with an output voltage so low. You cannot use voltage that low. I would strongly suspect that the reason for this is the simple fact that a higher voltage (i.e. putting more cells in series) would fry the life out of PV films that thin.
I have discovered what may be a problem with the thin-film PV mission concept.
As I stated previously, 100grams/m2 is the weight of individual cells tested in the lab, not the mass of the whole system. The thickness of vapour deposition amorphous silicon photoactive layer is ~1 micrometre. At this thickness, the breakdown voltage of the silicon is 30volts. Applying a reasonable safety factor, panel output voltage cannot much exceed 12volts.
For 10,000m2 of solar panels, divided into rolls 1m wide, I calculated the mass of copper cabling needed to transmit power from each roll of cells back to the edge of the array with no more than a 2V voltage drop. The answer is 113.613 tonnes.
I have not calculated the voltage drop across the panels themselves. Nor have I attempted to assess the energy loss due to the inverter and transformer, or the weight of these items.
For relatively small arrays deployed in space, where the transmission distance back to the inverter and transformer are just a few metres, the voltage drop is less of a problem. But larger systems like this would require that either additional mass budget is allowed for cabling, or that mass budget is allocated to local inverters and transformers which steps up the voltage for every 10m2 of cells. Another alternative would be to increase the thickness of the silicon layer and use a thicker backing material. But that adds mass to the panels.
My conclusion is that as a power source for a manned mars mission, thin-film PV is in its early concept stage and we do not yet have a good understanding of how a developed system would work. Planning around an assumed weight of the system is therefore premature.
1kg of oxygen is 13.1MJ of chemical energy, which requires about 20MJ of input to an electrolysis cell. That's 5.6kWh/kg. 5.6kWh divided by 24 hours is 233 watts continuous. Of course, if you are only powering that cell during peak sun, the actual power requirement is 1120 watts for 5 hours. And that cell must be larger, more massive and you will need to expend some energy heating it up each day before you can use it. Entropy is expensive.
The problem with titanium and other non-ferrous metals like aluminium and magnesium, is that they do not have an infinite stress cycle. Assuming you are talking about pressure vessels and not cryogenic storage tanks, there would be a finite number of pressure cycles before the tank became unsafe. Pressure vessels for repeated use are usually carbon steel or ductile low-alloy steel.
Flexible bladders are not a good idea. They would suffer low energy density due to the low density of the stored gas and at typical Martian temperatures, most elastomers are beneath their glass transition temperature. A steel pressure vessel pressurised to ~200bar is a better idea. Exactly what the original authors of your thin-film study recommended.
Also, there would appear to be little benefit to converting hydrogen into methane. Every new chemical transition you undertake wastes more energy. The more energy you waste the more panels needed, the more heavy cabling and transformer units required and the more EVA times needed to assemble all this stuff. Electrolysis is already an inefficient way of storing energy. Assuming you are electrolysing water to produce hydrogen and then burning the hydrogen in an engine, total efficiency is 25-30%. Adding methane synthesis to the set of reaction steps cuts that down to ~20%.
This is why Li-ion batteries are better. Most of what you put in, you get out the other end. They are heavier for sure, but realistically the solar power system will not be as light as 0.1kg/m2 when transmission is factored in. So efficiency is important in a mass optimized system.
I referenced SP-100 as a representative space nuclear reactor in the 2500MWth range. The project was cancelled in the 1990s when it became clear that NASA would not get funding for manned space flight projects beyond low Earth orbit. This made it redundant for the foreseeable future.
150W/Kg is the potential power density of the solar cells without encapsulation. It is not the power density of the whole system. It remains to be seen whether Megaflex can produce solar power systems in the 100KWe range that compete on a mass and performance basis with a space nuclear reactor in the same power range. These are cells that are literally as thin as paper.
I would be happy to see it happen, but I am sceptical. There are a lot of complications that aren't being accounted for in your simple scenarios. The need for encapsulation to avoid UV damage on Mars may be one. Maybe I am over-egging it. But I do know that radiation damage and oxidation are big enough problems here on Earth to require that all thin PV films be encapsulated in glass to prevent rapid degradation. On Mars, the oxidation problem is less severe, but the radiation damage rate is hundreds of times greater. How that will effect polymer lined PV cells is an unknown at present.
Another issue I can see is the breakdown voltage of silicon films that are literally microns thick. This limits output voltage of the cells to a few tens of volts. For solar arrays spanning several thousand square metres, you either need solar cells clustered around local inverters and transformers or you need some thick cables delivering power back to your hab or propellant plant. Either way, its more weight. Significantly more weight.
And then there is storage. As I have stated before, planning your mission around an intermittent power supply is likely to introduce significant problems. So you need storage. Li-ion batteries are more efficient than stored methane and oxygen, which reduces the area of cells needed and reduces both cell mass and surface deployment time. They are also a lot more reliable than combined electrolysis fuel cell system, which must include numerous pumps, heaters, membranes and heat exchangers. You were worried about the reliability of liquid metal cooled fission reactors with multiple independent power conversion loops. How reliable do you think a fuel cell will be in comparison? Batteries are more reliable and efficient, but they are heavy. Dust storms and dust accumulation are other issues.
Maybe it is possible to solve these sorts of problems and produce a reliable system in the 100KW range with a mass power density superior to a nuclear reactor. But wishful thinking won't get us there.
Apologies for not responding sooner. Between work and family, I get limited free time for hobbies like this. Louis and others requested information on the Specific power (KWe/Kg) of fast reactor power systems for deployment on Mars. Here is the information I have been able to find.
I found it difficult to track down detailed design concepts for the SAFE-400 reactor, as most such information is behind paywalls. I had more luck with Kilopower and SP-100. The original thin-film solar surface power concept (see below) referenced a specific power of 19W/kg for SP-100 with four Stirling cycle power converters delivering 100kWe. As the SP-100 is a 2.5MWth reactor system, these arrangements would require using the reactor at only 16% of its rated power, which seems less than efficient. The second concept was based on the Prometheus design for a lunar based reactor and had specific power of 10W/kg at power levels of 100kWe using a Brayton cycle.
http://systemarchitect.mit.edu/docs/cooper10.pdf
The specific power of space reactors appears to be a strong function of power output – the higher the power output of the system, the higher the specific power. This study presents a design arrangement for an SP-100 based system producing 550 kWe, with a total system mass of 11,879.9kg. That is a specific power of 46.3W/Kg.
https://www.osti.gov/scitech/servlets/purl/10181300
This study examine a lunar base power supply using the SP-100 concept and producing 825kWe, with total system mass 20,000kg – 41.25W/kg.
https://ntrs.nasa.gov/archive/nasa/casi … 005714.pdf
Conversely, for very small reactor systems, specific power is low, but increases progressively as power is scaled up (See Table 1 in the link, below). The Kilopower concept, achieves specific power of 2.5W/kg at 1KWe power level, but that increases to 6.5W/kg at 10KWe output.
https://ntrs.nasa.gov/archive/nasa/casi … 017750.pdf
Page 29 of the link below, provides a useful graphical approximation for system specific mass, for small fast reactors coupled to Brayton cycle generators. Both surface power concepts and space nuclear electric propulsion concepts are included.
https://ntrs.nasa.gov/archive/nasa/casi … 004957.pdf
• It can be seen that at power levels as low 10kWe, the specific mass is 350Kg/KWe, corresponding to specific power of 2.9W/Kg.
• As power scales up to 100KWe, specific mass is 50kg/KWe (20W/Kg). This compares very closely to what the researchers assumed in the thin-film solar study.
• As power requirements reach 1MWe, which is the power level required perhaps, for a large base, specific mass declines to ~18kg/KWe (55.6W/Kg).
• At power levels of 10-100MWe, what might reasonably be required for a colony of 1000-10,000 people, specific mass appears to converge towards 10kg/KWe (100W/Kg). Presumably at this point, radiator mass dominates the total system mass, as this part of the system obviously cannot realise any scale effects since heat ejected is proportional to radiator area. As reactor systems scale up, it makes more and more sense to attempt to build the radiators on Mars using local resources.
So, I would suggest that the best choice of power system for a Mars mission / base / colony, etc. is strongly dependant upon the power requirement (nuclear does much better as power scales up) and location, as solar power system mass would increase for locations further from the equator. There is the added complication that solar power systems including storage have their own specific power curves, with smaller systems having lower specific power than larger systems.
A large part of the power supply for a Manned Mars Mission does need to be available 24/7. Hotel requirements need to be maintained 24/7, as does power supply to communications and internal computer equipment. Heating/cooling the habitat must be carried out 24/7, although modest amounts of hot water can be stored. According to this source, these sorts of loads are anything up to 30kWe.
https://science.nasa.gov/science-news/s … spacepower
Other loads do not necessarily need to be 24/7, although that is always desirable.
•Cooling for propellant tanks might be allowed to vary if the tanks are well insulated and subcooled far beneath saturation pressure at the tank working pressure. But that reduces the COP of the coolers and may increase insulation mass.
•Vehicle recharging can take place during the day, but that tends to be when you would ideally wish to operate these vehicles.
•Science labs need not be operational full-time, but to get reasonable productivity from the crew, we would ideally need them to operational for 16 hours per day. Some processes may need to run continuously for days.
•Electrolysis can be carried out intermittently, but to get the same total mass of products, the electrolysis cells must be larger (and more massive) and must operate at higher power levels during operation. Efficiency is also affected, as these units work most efficiently at high temperatures.
•Propellant production need not take place continuously, but is more efficient if it does. The process involves electrolysis already mentioned. The Sabatier reactions take place at high temperatures and substantial amounts of energy will be consumed heating equipment up to operating temperatures. Compression of atmospheric CO2 is more efficient at night, as the feedstock is colder and denser and less compressor work is needed. To produce the same amount of propellant in the same amount of time, intermittent production will require a larger propellant plant, as the output per unit time is proportional to plant mass. High temperature equipment does not well tolerate thermal cycles. A propellant plant working on an intermittent power supply must be designed to tolerate hundreds if not thousands of thermal cycles without failure.
•ISRU processes like iron smelting are niche applications. But it is difficult to see why a day-time power excess would be an advantage. Real life high-temperature processes are not tolerant of thermal cycles and to achieve the best utilisation of invested capital tend to operate continuously on a steady-flow basis. Maybe you could demonstrate reaction constants for ore reduction at different temperature and CO/H2 gas flow rate using an intermittent supply, but that does simulate a real blast furnace. What happens if you need to leave the furnace running for 24 hours in order to achieve sufficient ore reduction? It isn’t much use if your peak solar power only lasts 4 hours.
•Greenhouses could be heated during the day and given sufficient thermal inertia to remain above freezing over night. But that requires additional mass such as water or phase change materials within the greenhouse to function as a thermal store. You could use dirt, but it has only half the heat capacity of water.
In short, there are no advantages to having an intermittent power supply. It can be tolerated, but tolerating it involves both mass and performance penalties elsewhere. For most things, it would be a lot more convenient to have a continuous 24/7 power supply. In order to properly assess the relative advantages of nuclear and solar we must either:
1.Compare them like for like, i.e. how much do they both weigh if we use them to produce a 24/7 power supply; or
2.Understand the mass and performance penalty of solar intermittency on the mission as a whole. We can then compare a solar mission to a nuclear mission as a whole and determine which is the best deal, when all costs and penalties are accounted for.
Since the second is much more difficult to quantify, most mission designers have thus far taken the first approach. That is the approach most of us have taken on this board. The second should be attempted for any mission proposal that is beyond concept stage.
Elon Musk is also promoting Solar technology through his Solar City business, as well as promoting the Tesla automobiles. Yes, he has the wherewithal to take us to Mars with his rockets, but so far has done little to explain what happens after we get there. He has yet to offer anything regarding infrastructure or continued support for those his company wants to transport there. Solar City is financially on the ropes, and Tesla is also fraught with difficulties.
I agree that Elon is a remarkable individual with a great deal of vision, but he too, can make mistakes.
Agreed. There is of course the inescapable fact that purchasing a small nuclear reactor would be an absolute administrative and legal headache for a privateer like Elon Musk. He may have decided that it was easier and cheaper to work with a more massive and less efficient energy system than to attempt to pick his way through the legal minefield of trying to procure and launch an SP-100. It is politics and legal skulduggery that has crippled the nuclear power industry here on Earth.
That said, the difference in mass between a small fast reactor and an encapsulated (UV protected) solar power system, both capable of delivering the same end use power output, is more than a factor ten. There is no way around that. Even on Earth, thin-film PV is typically encased between glass panels to prevent rapid UV degradation. Even then, efficiency declines at ~1%/year. Imagine taking thin-film solar cells into an environment with a UV damage rate hundreds of times worse than Earth’s with no encapsulation at all? That is essentially Louis’s scenario. That is before we get into issues like energy storage and deployment of the system on Mars.
In the long run, we will need to build power systems on Mars using energy and resources from Mars. The analysis in previous threads on this topic shows that that is unlikely to be energetically favourable to do that using solar power systems, with energy payback times of ten years or more.
These problems are a direct result of the poor energy density of sunlight on Mars. Sunlight is a time-averaged energy flux of about 140W/m2 on Mars. Compare that to a fast reactor core, which has power density of 300MW/m3. That's a difference of six orders of magnitude.
100% fact-free comment from Antius. Why do facts, when you can do ad hominem?
Antius wrote:Well Louis, you seem to be having fun. So I won't bore you with any inconvenient facts :-)
Louis, the problem is that we have gone over the facts, over and over again. Many hours of time were invested. We have done a lot of analysis in other threads and the thin-film based power supply has been shown to be highly problematic. So why start a new thread presenting it as a viable mission scenario, without answering any of the fundamental problems that have been uncovered? How many times do I need to kill this white elephant?
Well Louis, you seem to be having fun. So I won't bore you with any inconvenient facts :-)