The Space Reviewin association with SpaceNews

lunar landers
Three different versions of a reusable lunar ferry: cargo, tanker and crew, with a common propulsion section, will be much more useful than a single lander which has to be launched in three pieces due to politics. (credit: Anna Nesterova)

The struggle for a practical cislunar transportation system

Bookmark and Share

Although I have written several previous articles covering cislunar issues (see “Why use lunar propellant?” The Space Review, April 2, 2018), events have now clearly reached a whole new phase for all the players: within NASA, in the existing industry, and in NewSpace. This will strongly affect the central part of design and planning for beyond low Earth orbit (LEO) operations, the next major step in space development and operations by NASA and its international partners. This goes beyond the issue of using lunar propellant. Understanding the new situation and part of what may be driving it requires information about several different initiatives.

If that stage can be refilled with propellant from a Spaceship tanker, the first Starship will be able to go to the Moon, land, place at least 100 tons of cargo on the lunar surface, take off, and return to the Earth and land again. This could happen in as little as three years.

In late December, SpaceX released photos which show the assembly of a nine-meter-diameter stainless steel test vehicle at Boca Chica in far South Texas that will probably result in the first vertical test flights of a Big Falcon Spaceship (or under its odd generic new name, Starship) test vehicle in early 2019. Once the vehicle is flying and landing safely, and a spaceflight-capable stage has been built, it can be launched into orbit by a Bif Falcon Rocket booster stage (now called Super Heavy.) That means the path to a lunar demonstration trip by SpaceX will be open.

If that stage can be refilled with propellant from a Spaceship tanker, the first Starship will be able to go to the Moon, land, place at least 100 tons of cargo on the lunar surface, take off, and return to the Earth and land again. This could happen in as little as three years. Time seems to be running out for the SLS, since once the huge Super Heavy/Starship combination is operating for a tiny fraction of a billion dollars per flight and completing multiple launches and landings, it will become very hard to convince the Congress to continue support for the SLS, which will cost billions per flight and cannot be reused.

Meanwhile, players in the US other than NewSpace companies are starting to realize that in just a few years, they will probably be left in the dust by SpaceX, which seems no longer willing to slow down its own lunar trip capability to avoid embarrassing NASA. SpaceX is, of course, aiming at both the Moon and Mars. NASA’s timeline for putting human crews on the Moon again is not until 2028, a decade into the future, while SpaceX plans to send crews all the way to Mars in 2024 or 2026. In October, Lockheed Martin, which is definitely not a NewSpace company, published a perfectly reasonable and detailed design for a large reusable lunar lander or lunar “ferry,” a word which still seems to be prohibited within NASA. (A “lander” is a ferry that can only land, but cannot take off again!)

NASA has its own bizarre plans for a reusable lunar transport system, which in many ways is a good thing, but clearly still constrained by SLS politics. NASA had been focusing up until early 2018 on its seemingly useless Lunar Orbital Platform- Gateway or LOP-G, which some denigrate as a space station “Mini-Me”, or as a program simply to create something for the SLS rocket to launch. (Almost all of the funding which could have created a variety of large, capable payloads for rockets in the same general launch vehicle class as Falcon Heavy, New Glenn, SLS and the BFR, has been diverted to the SLS and the preceding Constellation program for over a decade, with the resulting lack of future-oriented NASA programs and hardware.) The much vaunted Mars program of the Obama Administration was a “Potemkin Program”, all PowerPoint, no hardware. Now that time is running out, NASA seems to be taking a new look at the critical cis-lunar transport problem.

NASA is officially unwilling to commit to launching any pieces of either a Lunar Orbital Platform-Gateway station, or an actual Gateway station with real propellant depot and logistical capabilities, with any rocket other than the SLS. It does not expect to be able to launch such payloads on the SLS until after 2026 or 2027, due to the slow pace of SLS development. However, it is willing to launch pieces of a supposedly reusable lunar vehicle on existing commercial rockets like the Falcon 9. NASA expects that a new set of lunar vehicles would be based at what it is now calling a Gateway station, with another cislunar set based at a LEO station. This indicates that opinion within NASA is now shifting to support a real gateway, instead of a fake one.

Including a real Gateway, logistics base, or waystation with a propellant depot into the transport system design means that the amount of delta-V needed to go from the Gateway to the lunar surface and back is about half of that needed to go from LEO to the lunar surface and back. Bringing that round-trip delta-V down to the range of about five kilometers per second or less makes it easy to design fully reusable vehicles, as the shorter trip “length” requires only about one fourth as much propellant for the same vehicle size as the longer trip. This is because the vehicle is not carrying all of the propellant for the other leg of the trip along with it. The huge and super-efficient Super Heavy/Starship system would not require a Gateway station for lunar missions due to its sheer size, but medium-size vehicles would need it so they could be more efficient.

Including a real Gateway, logistics base, or waystation with a propellant depot into the transport system design means that the amount of delta-V needed to go from the Gateway to the lunar surface and back is about half of that needed to go from LEO to the lunar surface and back.

To complicate the situation, there is the Lockheed Martin reusable lunar ferry design, which is also designed to be based at and operate from a Gateway station. It would either be delivered into LEO dry and refueled there or launched to LEO wet (with the propellant load inside). The first propellant load would take it from LEO to a lunar orbit or L1 gateway base, from where it could return to LEO via aerocapture with only five to six tons of propellant. This ferry is 22 tons dry and needs a second load of 40 tons of propellant delivered at the near-lunar Gateway base in order to reach the Moon’s surface and return, a 6.1 kilometers-per-second delta-V trip. Such a lunar ferry would be able to easily support a lunar base construction and operations, with a crew of four astronauts. A Falcon Heavy could easily deliver such a vehicle, partially fueled, into LEO, and a Falcon 9 tanker could deliver the rest of the propellant needed to reach the Gateway station.

However, NASA cannot not touch the Lockheed Martin design since they also refuse to launch a heavy lunar ferry on anything other than an SLS. That seems to be the primary reason they have now come up with a kluge of a lunar ferry, designed to be launched in three pieces. They can launch the pieces of the ferry years ahead of the Gateway station components, creating some good short-term PR. But since the “trinity” ferry cannot actually be used to land on the Moon and take off again until the Gateway station exists, this option is rather pointless.

The three sections of the three-part NASA lunar ferry are: the space tug section, the descent stage section and the ascent stage section. The first section is essentially a reusable space tug that would move the combination of the descent and ascent stage from the Gateway station in high lunar orbit (1,500 by 70,000 kilometers) down to low lunar orbit and then return to the Gateway by itself. The next section serves as the lander, similar to the old Apollo descent stage. It would carry the ascent stage from the Gateway down to the surface of the moon, but would stay there as an expended section. After a stay on the surface, the crew would eventually take off in the reusable ascent stage, which would take them all the way back to the Gateway.

The problem caused by this design with a trinity of ferry components is that the total round trip delta-V from a similar location, the L1 point, is only about 5.2 kilometers per second. The Lockheed Martin design clearly shows that a single-stage reusable vehicle like it would work fine, as its 40 tons of propellant would provide for a round trip starting from a lunar orbit with 4.67 kilometers per second of delta-V or a one-way amount of about 2.34 kilometers per second. Compared to a one-way trip to L1 of 2.6 kilometers per second, to reach the Gateway station’s elongated elliptical lunar orbit from the surface would take about 260 meters per second less delta-V. If the Gateway were placed at L1 instead of high lunar orbit, the vehicle would need just 8 tons more propellant for a round trip to the surface.

The other problems that the three-part lunar ferry would create include:

  1. The ascent stage would have no backup propulsion system to save the crew if the main ascent propulsion failed. This would put the crews in as much risk as the Apollo crews were during ascent.
  2. The need to reassemble the whole vehicle from its components, including one brand-new untested component—the descent stage—every time the vehicle was used, would be complicated by the small number of personnel to supervise a correct re-assembly, which would probably be performed by berthing rather than docking the parts together, and would probably require self-testing and reporting capability for all electrical and electronic connections between sections.
  3. The expendable descent stage, one of the largest components, would probably force the cost of each single use of the lunar ferry system up to a minimum of about $1 billion, as time has shown that no crew-carrying vehicle designed by NASA can cost less than about a billion each due to all of the safety-related processes the vehicle would go through. This alone would make the system nearly as fiscally unsustainable as launching missions with a new SLS each time.
  4. This design would make it much more difficult to create the alternative versions of the ferry, such as a crew cabin, flatbed cargo carrier, and tanker. All of these alternate sections should be using (and sitting on top of) an identical propulsion system section.

It is reasonable for NASA to create an alternative lunar transport system to the one SpaceX is building, as long as it is fiscally sustainable. We are currently making a great effort so we will not depend on a single transport system to reach low Earth orbit, as we are currently with Soyuz, so neither should we depend on a single system to reach a lunar base. So now, assuming that NASA planners would listen, what characteristics should have been (and could still be) laid down as the basis for a safe, practical, and fiscally sustainable lunar ferry and transport system design? From my point of view, they include:

  1. The whole ferry vehicle should be reusable and able to return to its base, either in space or on the lunar surface, without refueling.
  2. The vehicle should not separate into parts except during a dire emergency where there is loss of vehicle control, loss of thrust, or other severe damage.
  3. The cargo version should be able to deliver at least 20 to 25 tons of one-way cargo (the mass of a lunar crew habitat) to the lunar surface and off-load it (in cargo or tanker mode), and the design should support a version with a lifeboat-capable crew cabin of about seven to nine tons including emergency propellant, as the cabin and propellant mass is carried in both directions on a round trip.
  4. The crew cabin must have its own separate emergency propulsion to save the crew if the main ferry fails in flight, able to generate at least 1.3 kilometers per second of delta-V, enough to do either an abort to lunar orbit or abort to the lunar surface. For a nine-ton crew cabin, that would take just three tons of emergency propellant.
  5. If larger cargo masses are needed than the medium-sized system described, the ferry system design should be scalable (before fabrication starts) to the optimal or largest cargo or propellant mass payload anticipated, such as a large lunar crew habitat or mining vehicle.
  6. The three ferry versions should have a common propulsion section to save costs (in the bottom position of a lander stage), with one of three different sections on top of it (crew cabin, tanker, flatbed cargo carrier.) A crew ferry cannot carry large cargo items, and a cargo ferry cannot carry a load of propellant without propellant tanks.
  7. The crew cabin should carry at least three to four people with full life support for a month.
  8. The vehicle must be able to withstand a lunar night without damage (external thermal blanket covering is a possibility.)
  9. All vehicles should use liquid oxygen and liquid hydrogen propellants for maximum efficiency and should have cryo-coolers and insulated, shaded tanks to preserve the propellant from boil-off. As described in my earlier article, such a system is extremely efficient when moving propellant to lunar orbit or cargo to the lunar surface.
  10. Since a transport system must include both vehicles and (stationary) nodes, it must have a refueling depot and logistics base either at the Earth-Moon L1 or L2 points or in a near lunar orbit, with the vehicle designs integrated with the base design and location. (The refueling base should also enable the transfer of cargo from the cislunar ferries to either lunar or Mars-bound vehicles and should later start accumulating propellant for the Mars expeditions, tying the Moon and Mars projects together.) The Gateway station can be such a base and gradually expand as needed.
  11. The transport system should include a corresponding set of cislunar ferries similar in design to the lunar ferries, but operating between LEO and L1, L2, or a lunar orbit. The main difference is that they would have thermal protection and an aerodynamic exterior for returning to LEO via aerocapture instead of landing legs.
  12. Any lunar program should have a primary initial objective of creating a lunar polar base for mining volatiles with water ice, with its main initial function being to set up a lunar propellant production system, so that the propellant can then be supplied to vehicles in LEO or headed for Mars or asteroids.

The lunar ferries would be initially supplied with propellant at the Gateway station, supplied by cislunar tankers coming from LEO. When the crew and cargo ferries are operating between a Gateway station and the lunar surface, the tanker ferries would not get much use, as each ferry carries its own return-to-orbit propellant that it got at the Gateway. A reserve of propellant should be maintained at the lunar base from the start.

Some even think that NASA is planning a bait and switch game with the Gateway and do not plan to land on the Moon.

However, once lunar propellant was being produced, the lunar cargo ferries would be able to bring down larger loads, as they would be refueled on the surface instead of at the Gateway, and the lunar tankers would then be put into frequent service to carry large amounts of propellant to the Gateway, wherever it is located. This lunar propellant could then also be used to supply the cislunar ferries operating between LEO and the Gateway, which would result in a large drop in operational costs.

The Lockheed Martin large reusable ferry design probably comes closest to meeting the basic requirements above. However, while it does seem to include a logistics base, the current NASA “trinity” lander concept does not meet most of these specifications, and it is also a program with a very slow schedule. Some prominent people in our community clearly recognize that. Other observers are still confused about what NASA is actually proposing, and if it will be of any use for building a lunar mining base. Some Mars proponents do support the Gateway station concept, since it could develop into a major transportation hub which could support Mars expeditions with cargo handling and propellants. They wish it were at L1 instead of in lunar orbit, since L1 is a much better place to fuel up from.

There are others with proposed lunar architectures, such as Robert Zubrin, who do not want any kind of space infrastructure, including an actual Gateway station. They think that somehow it is a distraction from the main program, which is getting to a “real” destination like the surface of the Moon or Mars. Some of them even think that NASA is planning a bait and switch game with the Gateway and do not plan to land on the Moon. (For this to be a believable plot, a huge number of NASA employees would need to be in on it, which strais credulity.)

Zubrin recently issued a fairly detailed paper on his proposed architecture called Moon Direct. In this paper, one vehicle called the “cargo lander” carries the crew ascent vehicle all the way from LEO to the lunar surface instead of just from the Gateway station, and then the crew ascent vehicle, or LEV, has to go all the way back to LEO, not just to lunar orbit. That is 6.1 kilometers per second each way, a total of 12.2 kilometers per second delta-V for the round trip, instead of just 5.2 kilometers per second or less for a round trip to the Gateway. As a result, the vehicles need much more propellant per ton of dry mass, and the allowable dry mass is very tiny.

Zubrin’s cargo lander vehicles are so small that they could not deliver much round-trip cargo to the lunar surface per mission (an eight-ton crew habitat is pretty small), and the lander would be expendable, like NASA’s. They would probably mass between 56 and 62 tons with their eight-ton payloads of LEV or cargo. The LEV vehicle itself is the same size as the Apollo lunar ascent stage—it has only two tons of dry mass but needs three times more fuel—and it might be able to hold one or two crew members at best. (Zubrin does not seem to say what the crew size is). There would be essentially no shielding mass available to provide solar radiation protection for the crew during the trip to the Moon and back.

The Moon Direct concept is thus another minimalist plan that would simply not accomplish much or leave much permanent infrastructure at the lunar base site, and it greatly underestimates the tonnage of equipment needed at the mining base site for doing anything practical. Zubrin’s design is primarily intended to support lunar surface science exploration, and not to supply lunar propellant for other uses to a propellant depot in space. As described in his paper, the design fails to meet the above criteria in multiple ways.

I truly wish that NASA would be allowed to just go ahead and design and build a practical lunar transport system, using whatever commercial launchers will be available in the next four years, including the Falcon Heavy, the New Glenn and Super Heavy/Starhip, to launch both intact, full-size lunar ferries and the Gateway station sections into orbit as well as get a lunar water mining program rolling. In the future, there will be lots of traffic leaving the Earth-Moon system, and the L1 and near-lunar zone is one of the best and most efficient departure points. Both Moon and Mars proponents should support a real Gateway, and real reusable ferries.

Note: we are temporarily moderating all comments subcommitted to deal with a surge in spam.