With Elon Musk, of SpaceX, about to announce his master plan to send mankind to Mars (using a massive Super Heavy-lift Launch Vehicle) and with unmanned SpaceX Dragon 2 flights to the planet already in the pipeline, the question that many are asking is: was NASA’s development of its own SLS Heavy-lift Launch Vehicle (HLV), a waste of time and money?

Artist's impression of SLS Block 1 ready to launch. Courtesy: NASA

Artist’s impression of SLS Block 1 ready to launch. Courtesy: NASA

 

While the in-line configuration SLS project has gone well, it has proved very expensive with a projected development cost of US$18 billion (until 2017).  Worse than this, however, is that NASA has very few missions to fly on it. This has led some critics to suggest that NASA has built a rocket with nowhere to go, and one that could soon be made redundant by other launch vehicles.

Yes – NASA’s rocket is too expensive but at least it now has a basic heavy lifter

After the Augustine Commission’s findings ended Project Constellation, and the development failures of Ares 1 and Ares V, NASA found itself in want of a large rocket for crew exploration. Or rather it and the Whitehouse wanted one, but only after it had fixed up commercial LEO spaceflight first. However, their hand was forced by the US Senate, in cahoots with former Apollo astronauts such as Gene Cernan and the now late Neil Armstrong, and so a commitment was made to a Heavy-lift launch Vehicle (HLV) using elements of the Space Shuttle programme.

Considering the design in 2010-11, NASA originally favoured a much cheaper (US$6-8 billion), if less capable, Sidemount HLV configuration. This derivative of the Shuttle-C concept would have made use of most of the Shuttle’s hardware and launch infrastructure. However, after comparing different configurations, NASA instead chose the much more expensive in-line SLS (Space Launch System) design. It based this decision on the grounds that SLS had greater development potential (true) and that Sidemount had complications in rocket-powered crew escape scenarios (not entirely true).

The Sidemount HLV could have been built for one third of the cost of SLS. Courtesy: NASA

The Sidemount HLV could have been built for one third of the cost of SLS. Courtesy: NASA

 

One wonders whether the disputed crew escape issue was a red herring deliberately deployed by the in-line side. There were concerns, but analysis showed that Sidemount crew escapes would have worked – see HLV Crew Abort Assessment. If there had been a problem, Sidemount could have just been used for the unmanned portion of any two-launch mission, even if this might have meant the revival of the Ares I to carry Orion.

Opponents of the entire plan to develop any HLV claimed that NASA simply did not need one. Instead they suggested it should rely on new commercial heavy lifters, which were likely to become available (i.e. the Falcon Heavy) and/or use lots of smaller launch vehicles in combination with new in-orbit technologies, such as L2-located fuel depots to mount exploratory missions. Neither of these imaginative suggestions was actually practical.

In fact, large as it is, the 53 metric ton class Falcon Heavy (Falcon 9 Heavy) is too small even for basic lunar missions. Meanwhile, the multi-launch/fuel depot plan relied too much on untried or undeveloped technology. It would also have required a long-term political commitment to fund a great many launches – something that cannot be depended upon given the fickle and cyclical nature of democratic politics.

Therefore, an HLV was definitely needed. The trick that NASA needed to pull off was to produce one as cheaply as possible.

And it was possible. As this column pointed out (Space Intelligence News – September 2009), the 80-90 metric ton payload Sidemount HLV offered an adequate lifting capability for about one third of the cost of the in-line option, and would thus have been a much more cost-effective choice as an “Interim HLV” for NASA.

Importantly, it would have been much quicker to develop than an in-line like SLS; it could have been used at the same time as the Space Shuttle, or at least experienced a smooth transition from it; and it would have used exactly the same launch pads, construction and support infrastructure.

Choosing Sidemount would also have had the advantage that its much lower cost would have freed up funds for other exploration hardware – a lunar lander for instance.

While the premature demise of the Space Shuttle programme took away some of the main advantages that Sidemount had, it still remained the best “Interim HLV” choice.

At this point, it has to be conceded that there was some logic in NASA’s and the US Senate’s decision to go for the in-line SLS design. When NASA chose it, the SpaceX Mars Colonial Transport/SHLV combination was still just a twinkle in Elon Musk’s eye. So, they picked a design that might be turned into a more powerful HLV for later Mars missions. However, as time has passed, this choice increasingly looked like the wrong one.

Sidemount should have been a lunar workhorse but SLS Block 1 could take up the reins

Sidemount would have given NASA the power to mount manned expeditions to the Moon again. The Augustine Committee heard from NASA’s then Space Shuttle Programme Manager, John Shannon, that Sidemount could and should have been developed for manned lunar missions using two launches.

Two launch mission plan using the Sidemount. Courtesy: NASA

Two launch mission plan using the Sidemount. Courtesy: NASA

 

This column agreed with this suggestion, adding that, a much larger (200 metric ton plus payload) rocket – a Super Heavy-lift Launch Vehicle (SHLV) – could be developed for later Mars missions.

This is now effectively happening but, instead of the Sidemount, it is the in-line SLS that is likely to become the “interim HLV” – and a very expensive one at that. Meanwhile, SpaceX is developing an SHLV to carry the “Mars Colonial Transport”, which is likely to make later versions of the SLS redundant.

In effect, NASA “chose the wrong runner” when it went for an expensive thoroughbred like the SLS instead of a simple and much cheaper workhorse in the Sidemount. The latter would have got them from A (Earth) to B (the Moon) for a much cheaper price and in a much shorter development timeframe. Meanwhile SpaceX is now breeding a giant Shire horse, in the form of a (probably part-reusable) SHLV, to carry its very heavy loads all the way to C (Mars), making higher performing versions of SLS obsolete.

And since NASA could only afford to build SLS but not key payloads, save for the Orion itself, SLS will now become an under-used “rocket to nowhere”. Worse, it seems that NASA always knew this since its Michoud facility is only able to knock out, at best, two complete SLS vehicles per year.

Hindsight is a wonderful thing.

Still, while the Sidemount horse has now bolted, and while NASA did spend US$10 billion too much, nevertheless, at least it does now have some heavy-lift launch capacity in the form of the SLS Block 1 at its disposal.

NASA now has to show both the critics of HLVs and the US taxpayer that investing in SLS was worthwhile. It can only do this by actually using it – and as quickly and as much as possible. That means using the SLS Block 1.

Current plans mean that SLS will be under-utilised

NASA’s limited funds have forced it to build the SLS slowly at the expense of other payloads, apart from Orion. Combined with its reluctance to do proper lunar exploration, this means there is currently a dearth of missions for the SLS launch vehicle. As it is, even the two that are fully planned are delayed.

The first of these launches (EM-1) will now take place in late 2018 (originally set for 2017). The EM-1 mission will use the SLS Block 1 fitted with an Interim Cryogenic Propulsion Stage (ICPS), derived from the Delta IV rocket’s upper stage, to send an unmanned Orion capsule and service module combination around the Moon. The ICPS is powered by a single Aerojet Rocketdyne liquid hydrogen/liquid oxygen (LOx) burning RL-10B-2 upper-stage rocket engine.

 

Exploded view diagram of SLS Block 1. Courtesy: NASA

Exploded view diagram of SLS Block 1. Courtesy: NASA

The second flight (EM-2) will be manned but will not take place until 2023 (originally planned for 2021). This is because, under US congress diktat, the flight has to use the intermediate 105 metric ton payload SLS Block 1B rather than the already flying SLS Block 1. The Block 1B will employ the yet-to-be-developed man-rated cryogenic Exploration Upper Stage (EUS), hence its lateness. The EUS design is now expected to use four Aerojet Rocketdyne liquid RL-10C-1 upper-stage engines, beating off competition from Europe’s Vinci design.

Under current (and much vaguer) plans, the Block 1B rocket, regarded as the “workhorse”, will eventually be used for a manned flight to a captured asteroid pre-positioned in cis-lunar space. This is supposedly a more relevant test of the technology needed for Mars exploration missions.

Orion capsule/service module combination with ICPS transfer stage attached. Courtesy: NASA

Orion capsule/service module combination with ICPS transfer stage attached. Courtesy: NASA

 

NASA’s idea is eventually to fit the Block 1B version with advanced boosters (either solid or more likely liquid fuelled) to covert it into the Block 2, which should be able to carry 165 tons to low Earth orbit and send more than 45 metric tons on a Trans Lunar Injection (TLI) trajectory. However, this rocket will not be available until the late 2020s.

With only two test flights definitely procured, SLS is already too under-utilised to make the funds and time spent on it worthwhile. SLS launches are officially estimated as costing US$500 million each, but if development costs are taken into account, the two flights come in at a massive US$9 billion per flight. If the Orion costs are included, the launch cost breaches US$12 billion, making them easily the most expensive launches ever to fly. Perhaps it is unfair to allocate the development costs in this way, but even if only the annual SLS operational costs are allocated to these launches, the bill is still likely to breach US$1 billion a pop.

NASA should use SLS Block 1 for two-launch landing missions on Moon

While SLS could be be used in the construction phase of a Mars mission, even the most powerful version does not have quite enough payload capacity and definitely does not have a fast enough flight rate at only one or two flights per year. According to past NASA Mars exploration studies, a minimum LEO payload of 200 metric tons was suggested with mission construction launches being made in quick succession. As such, the SHLV role should now be left to SpaceX.

However, SLS could still have a significant role in manned lunar exploration, for which it is an ideal sized, but it needs NASA (and the US Congress) to choose to go down this path. If NASA did decide to do this, it could rapidly up the SLS flight rate. However, until now, it has not been keen to invest its limited funds in lunar exploration architecture because this would drain resources from its long-term Mars effort. Instead NASA wanted to concentrate on missions that might be more applicable to research for (much) later Mars missions.

A growing faction within the space community is now demanding a change of course and destination. These critics have realised that NASA’s Mars plan is too vague and too long term to be viable. Instead they suggest that the Moon is a better choice of destination and that NASA already has most of the elements it needs to make the journey.

NASA would not even have to wait for later versions of the SLS.  The nearly ready Block 1 is capable of Sidemount-like performance in being able to carry 86 metric tons to LEO and 25 metric tons to CIS lunar space (despite its 70 metric ton performance sticker), As such, the Block 1’s ICPS might be man rated with extra meteoroid shielding, which would allow it to be used at least for initial sortie missions to the Moon. If the ICPS had extended tanks (a relatively cheap modification), the Block 1 might reach well above a 90 metric ton payload to LEO and circa 35 metric tons to TLI – the sort of numbers first envisaged for two-launch lunar landing architecture using the Sidemount Block II.  Others have thought of an even cheaper solution.  By simply adding a Centaur upper stage to the ICPS, the rocket becomes capable of lifting 40 metric tons to TLI.

Alan Bean climbs down ladder onto lunar surface during Apollo 12. Courtesy: NASA

Alan Bean climbs down ladder onto lunar surface during Apollo 12. Courtesy: NASA

 

By using a pair of Block 1 rockets equipped for a dual-launch mission, it would be possible to mount a basic lunar landing mission in a relatively short time. While two-launch architecture is usually more expensive on a per mission basis, overall costs compared with building a single super-large SLS will be much less given the low number of missions involved. While a single pad (Pad 39B) could be used for two successive flights, ideally a second pad would be used – perhaps it is time to get back Pad 39A from SpaceX.

NASA knew that if the first manned flight of SLS was going to be in 2021, it would have to use a man-rated ICPS on a Block 1 rocket. But the US Congress demanded that NASA halt work on man-rating the ICPS in February this year. It may have sensed the existential danger to more powerful models of SLS, albeit while having more genuine concern about the ICPS’s resistance to meteoroid strikes.

NASA was instead ordered to redirect funds towards developing the more powerful and rugged Exploration Upper Stage (EUS) needed to power the later Block 1B and Block 2 versions. This political interference caused the delay to 2023 in the first manned SLS/Orion flight.

By the way, while the Block 2 version would theoretically allow NASA to perform cheaper Apollo-style single launch missions, with a small manned landing craft, it will not be ready for decades. And by the time the Block 2 is ready, the Space X super rocket (dubbed BFR – Big F***ing Rocket) may have already usurped it.

It is true that using higher performance liquid propellant boosters to replace the current dirty and inefficient solid rocket boosters (SRB) on SLS is an attractive option. This could be done by using either F-1B powered expendable designs or existing Falcon 9FT first-stage rocket stages as boosters. But these are not needed for now.  In any case, employing reusable boosters to try to reduce operational costs would probably not work given the low flight rates and large performance penalty.

Meanwhile, the EUS equipped Block 1B, which NASA now plans for the first manned flight, would still not be powerful enough to carry a small lunar lander as well as a full Orion craft, and so would need two launches anyway. In fact, there are major doubts about whether the intermediate Block 1B version is worth doing, given its relatively modest extra uplift compared with a Block 1 with an extended tank ICPS.

The problem here is that if NASA cancelled the EUS and Block 1B and used the Block 1 instead, it will be tacitly admitting that it should have built Sidemount in the first place. This it probably will not do.

A two-launch Moon-landing plan could be made to work with a Block 1 if its ICPS was man rated and its tanks extended. However, an Apollo-style lunar landing craft would also need to be built.

The mission would be launched first by a Block 1, followed a few days later by another Block 1 carrying astronauts in an Orion spacecraft. The Orion/Service Module package would probably meet up with the lander using a lunar orbit rendez-vous technique. With some of the astronauts aboard, the landing craft would then split off to reach the lunar surface. After some days, its ascend stage would lift off to rejoin Orion and complete the return trip to Earth, with an Apollo-style splashdown a few days later.

NASA should cut EUS and concentrate on using SLS Block 1 version with a lunar lander

NASA’s HLV development programme should always have been about making the best of what it has. While some US Senators followed “pork barrel” political self interest in wanting to keep NASA work in their states, their insistence that existing Space Shuttle hardware should be used for the new HLV was the right thing to do. Things only went wrong when NASA strayed too far from this path and chose the SLS over Sidemount.

The way forward is for NASA to cut EUS and any future versions of SLS and use the nearly completed SLS Block 1 as soon and as much as possible for lunar exploratory flights.

NASA should ensure that the ICPS is fully man rated/meteoroid protected, while producing a minimal change (and low cost) extended-tank version to maximise TLI payload. It should also develop a small expendable lunar lander for basic surface exploration and part-finance SpaceX to develop its SHLV rocket for Mars. By doing this, the first manned flight of Orion, EM-2, would be able to fly in 2021. EM-3 – a dress rehearsal for a landing flight – would take place in 2022 with EM-4.

After basic exploration of the lunar surface, and assuming the Mars flight has not occurred by then, proper lunar architectures could then be developed. This would include a small space station, such as the proposed Skylab II, located in either lunar orbit or in a cis-lunar L2 position halo orbit. Such a station could eventually have refuelling facilities/fuel depot for a later reusable lunar lander, with a separate space transportation stage for trips to and from lunar orbit.

Perhaps, even a lunar surface base or “lunar village”, as mooted by Jan Woerner, head of European Space Agency (ESA), might be constructed eventually.

Others have had similar ideas: here is Marcel Williams’ concept for a similar reusable lander mission. Boeing has suggested similar plans here.

While the SLS Block 1 would be under-sized for injecting very large modules into TLI, electric propulsion could be used for a slower ride to boost larger elements from LEO to the cis-lunar environment. In fact, it is only the manned elements and those parts suffering from propellant boil-off that would need fast passages.

A very large rocket may eventually be needed for the quick and economic construction of a Mars mission. Thus SLS would still be in NASA’s back pocket, ready – in the event that the SpaceX super rocket does not happen – to be developed into the powerful SLS Block 2 version, with the EUS and new advanced (probably liquid fuel) boosters. But for now, large expensive rockets such as the SLS Block 2 and even the SLS Block 1B are simply not required.

Summary: NASA should back SpaceX Mars plan and use SLS Block 1 to build lunar infrastructure instead

Funding remains tight, but if NASA (and the US Congress) decided to give up its own Mars plan and support the SpaceX effort instead, it could afford to put US astronauts on the Moon in the meantime. This could be rapidly achieved by using the already available SLS Block 1, developing a man-rated ICPS with larger fuel tanks and building a new lunar landing craft.

However, before any of this can be done, NASA, and the US Senate and Congress, will have to learn how to swallow their pride.

Update on 20 May 2016:  Since this opinion piece was written Orbital ATK has announced their design for a small space station to be in place in CIS Lunar space.  This would be built using modules based on their Cygnus craft, which would be launched by commercial vehicles – probably the Atlas V. The idea is that this pre-positioned space station would be a target for astronauts flying on SLS flight EM-2 to visit.

Lunar orbiting station Courtesy: Orbital ATK

Lunar orbiting station Courtesy: Orbital ATK