Audit Finds Significant Challenges for NASA’s Journey to Mars

The LEMNOS project will provide laser communications services to NASA’s Orion vehicle, show in this artist concept. (Credit: NASA)

A new audit by NASA’s Office of Inspector General has found the space agency faces significant obstacles in its plan to send astronauts to Mars in the 2030’s.

NASA’s initial exploration missions on its Journey to Mars – EM-1 and EM-2 – face multiple cost and technical challenges that likely will affect their planned launch dates,” the report states about missions planned for 2018 and 2021.

“Beyond EM-2, NASA’s plans for achieving a crewed Mars surface mission in the late 2030s or early 2040s remain understandably high level, serving as more of a strategic framework than a detailed operational plan,” the report added.

The report’s findings include:

  • NASA’s expenditures on the Space Launch System (SLS), Orion and Ground Systems Development and Operations (GSDO) will reach approximately $23 billion by the end of fiscal year 2018
  • NASA’s budget projections for human exploration to Mars exceed $210 billion by 2033
  • SLS, Orion and GSDO programs’ average monetary reserves for the years leading up to EM-1 are much lower than the 10 to 30 percent recommended by Marshall Space Flight Center guidance
  • software development and verification efforts for all three programs are behind schedule to meet a November 2018 EM-1 launch
  • NASA does not have a life-cycle cost estimate or integrated schedule for EM-2, which makes it difficult to understand the full costs of EM-2 or gauge launch date assumptions
  • NASA’s plans for achieving a crewed Mars surface mission in the late 2030s or early 2040s are more of a strategic framework than a detailed operational plan
  • Journey to Mars strategy does not identify key system requirements other than SLS, Orion, and GSD
  • Strategy does not offer target mission dates for a crewed orbit of Mars or landings on the planet’s surface or nearby moon
  • Significant development work on key systems such as a deep space habitat, in-space transportation, and Mars landing and ascent vehicles must be undertaken in the 2020s
  • NASA has explored reusing systems and subsystems, developing new acquisition strategies, and exploiting technology innovations to help reduce the high cost of deep space exploration.

NASA OIG made six recommendations to address the concerns, including implementing a master schedule for the EM-2 mission, developing more rigorous cost and schedule estimates for SLS and GSDO infrastructure for EM-2, and pursuing a strategy to collaborate with other space agencies.

The space agency’s management concurred or partially concurred with the recommendations. NASA OIG said the agency’s proposed corrective actions were responsive except for establishing more rigorous cost and schedule estimates for SLS and GSDO infrastructure for EM-2.

The audit’s summary follows.

NASA’S Plans for Human Exploration Beyond Earth Low Earth Orbit
NASA Office of Inspector General
Report No. IG-17-017
April 13, 2017

Why We Performed This Audit

Human exploration of Mars has been a long-term goal of NASA and the Nation for the past 5 decades. In 2015, the Agency announced its Journey to Mars framework for deep space exploration with manned missions to Mars beginning in the 2030s. In addition to the technical and health-related challenges of deep space missions, such a multi-decadal venture will be very expensive, with NASA’s budget projections for human exploration to Mars exceeding $210 billion by 2033.

A vital part of achieving its long-term human exploration goals is the successful development of NASA’s new spaceflight system – the heavy-lift Space Launch System (SLS) rocket, the Orion Multi-Purpose Crew Vehicle (Orion) capsule, and the ground processing and launch facilities (Ground Systems Development and Operations or GSDO) needed to launch the rocket and capsule for cislunar and deep space exploration. NASA has invested more than $15 billion in these three programs since 2012, and its near-term goals include a first uncrewed flight of the integrated SLS/Orion systems –Exploration Mission-1 (EM-1) – no later than November 2018 followed by a crewed flight – Exploration Mission-2 (EM-2) – as early as 2021.1 However, NASA’s plans beyond these two missions are less clear, with several options in early development, including robotic and crewed missions to an asteroid in the early to mid-2020s to test technologies and capabilities that would be needed for a mission to Mars. Moreover, these scenarios were developed during the previous administration, and the Agency’s new leadership is seeking to modify those plans with the President’s fiscal year 2018 budget request proposing cancellation of the Asteroid Redirect Mission and the Agency issuing a document in March 2017 that modifies and fleshes out some of its plans.

In light of the enormous costs and challenges and the critical decisions that must be made in the next several years, we examined NASA’s plans for human exploration beyond low Earth orbit in the near-term, mid-term, and long-term. Specifically, we assessed the Agency’s (1) plans for and progress towards its first flights of the integrated SLS/Orion systems in the next 2 to 5 years, (2) challenges in executing a sustainable and affordable plan to send a crewed mission to Mars in the 2030s or 2040s, and (3) strategies to help reduce the costs of its human exploration efforts. To complete this work, we analyzed cost data, interviewed Agency officials, conducted on-site inspections, and reviewed planning documents, feasibility studies, and other relevant program documentation.

What We Found

NASA’s initial exploration missions on its Journey to Mars – EM-1 and EM-2 – face multiple cost and technical challenges that likely will affect their planned launch dates. Moreover, although the Agency’s combined investment for development of the SLS, Orion, and GSDO programs will reach approximately $23 billion by the end of fiscal year 2018, the programs’ average monetary reserves for the years leading up to EM-1 are much lower than the 10 to 30 percent recommended by Marshall Space Flight Center guidance. Low monetary reserves limit the programs’ flexibility to cover increased costs or delays resulting from unexpected design complexity, incomplete requirements, or technology uncertainties. Moreover, software development and verification efforts for all three programs are behind schedule to meet a November 2018 EM-1 launch. Finally, NASA does not have a life-cycle cost estimate or integrated schedule for EM-2, which makes it difficult for Agency officials and external stakeholders to understand the full costs of EM-2 or gauge the validity of launch date assumptions.

Beyond EM-2, NASA’s plans for achieving a crewed Mars surface mission in the late 2030s or early 2040s remain understandably high level, serving as more of a strategic framework than a detailed operational plan. For example, the Agency’s current Journey to Mars strategy does not identify key system requirements other than SLS, Orion, and GSDO, or offer target mission dates for a crewed orbit of Mars or landings on the planet’s surface or nearby moon. If the Agency is to reach its goal of sending humans to the vicinity of Mars in the 2030s, significant development work on key systems such as a deep space habitat, in-space transportation, and Mars landing and ascent vehicles must be undertaken in the 2020s, and the Agency will need to make these and many other decisions in the next 5 years or so for that to happen. In addition, to position itself to make wise investment decisions, NASA will need to begin developing more detailed cost estimates for its Mars exploration program after EM-2. More concrete estimates will also be necessary as Agency officials work with Congress and other stakeholders to ensure the commitment exists to fund a mission of this magnitude over the next several decades. In addition, NASA’s decision whether to continue spending $3 to $4 billion annually to maintain the International Space Station after 2024 will affect its funding profile for human exploration efforts in the 2020s, and therefore has implications for the Agency’s Mars plans.

NASA acknowledges that to successfully execute the Journey to Mars, cost saving measures and cost sharing must be part of its strategy. Consequently, the Agency has explored reusing systems and subsystems, developing new acquisition strategies, and exploiting technology innovations to help reduce the high cost of deep space exploration. In addition, sharing costs with foreign space agencies and the private sector could help NASA reduce its overall costs, and NASA is partnering with industry to conduct multiple trade studies on the systems needed for the Journey to Mars and providing technical and mission support to Space Exploration Technologies Corporation (SpaceX) related to the company’s planned uncrewed Mars mission. Moreover, the recently enacted NASA Transition Authorization Act of 2017 cites expanding permanent human presence beyond low Earth orbit together with international, academic, and industry partners as the country’s long-term goal for human space exploration efforts.

What We Recommended

To increase the fidelity, accountability, and transparency of NASA’s human exploration goals beyond low Earth orbit, we recommended the Associate Administrator for Human Exploration and Operations (1) complete an integrated master schedule for the SLS, Orion, and GSDO programs for the EM-2 mission; (2) establish more rigorous cost and schedule estimates for the SLS and associated GSDO infrastructure for EM-2; (3) establish objectives, need-by dates for key systems, and phase transition mission dates to flesh out its Journey to Mars framework; and (4) include cost as a factor in NASA’s Journey to Mars feasibility studies when assessing various potential missions and systems. To improve cost savings efforts, we recommended the Associate Administrator for Human Exploration and Operations (5) design a strategy for collaborating with international space agencies in their cislunar space exploration efforts with a focus on advancing key systems and capabilities needed for Mars exploration, and (6) incorporate into analyses of space flight system architectures the potential for utilization of private launch vehicles for transportation of payloads.

We provided a draft of this report to NASA management who concurred or partially concurred with our recommendations and described planned corrective actions. We consider the proposed actions responsive to all but recommendation 2, and therefore will close those recommendations upon verification and completion of the actions. For the remaining open recommendation, we will continue to work with the Agency to resolve our concerns regarding establishing more rigorous cost and schedule estimates for the SLS and associated GSDO infrastructure for EM-2.

1 In February 2017, the Acting NASA Administrator instructed the head of NASA’s Human Exploration and Operations Mission Directorate to study the feasibility – from a cost, safety, and technical standpoint – of adding crew to the EM-1 mission.

Save

Save

Save

  • ReSpaceAge

    Conclusion
    NASA’s Journey to Mars is Bull$#!t

  • delphinus100

    To think (according to this) they might get there when I’m about 90, and there was a time when I was pretty sure they’d get there before I was 30…

  • Charles Lurio

    Ditto. I thought the 1986 expedition would be too early for me to get onto…

    Now I just hope to be alive and mentally coherent when _somebody_ goes.

  • redneck

    One of the ironies that we learned from Obama is that it is difficult to eliminate this waste without congressional (Shelby etc) cooperation. Hopefully a better approach emerges from the current disaster of a hashtag program.

  • Kapitalist

    I wonder if such a very damaging report being released indicates a change in policy?

    I also wonder how the politically correct space lobbyists, like Bill Nye, will react to this. It is after all NASA saying this. Will they obey their new master as thoughtlessly as they obey their last one? Or will they be annihilated when this hits them like matter hits anti-matter?

    I’m disappointed that they don’t conclude that Orion should be scrapped immediately. It is completely useless. Given all the sunk costs, I can somewhat understand trying to do something out of the SLS. It can after all lift some stuff to orbit. But it would be very stupid and dangerous to launch humans on it.

  • Obediah Headstrong

    I somehow get the notion that a manned trip to Mars will not be undertaken as long as the present political constellation is at hand. And possibly it will always be something just out of reach because…

  • JamesG

    … that is the way that they want it.

  • JamesG

    Its still excessively expensive to operate it. Nothing short of some national imperative (aliuns!) to get production and flight rates up to a dozen or more per year will even put a dent in the inefficiency and cost of it. It was a dead end from the moment it was proposed.

  • JamesG

    Well, mostly its the supporting documentation for scrapping the previous administration’s “plan” and funding priorities. IOW, what always happens. But I’m not optimistic that common sense and rational though is going to break out at NASA management any time soon. Congress is driving this clown car.

  • ReSpaceAge

    SpaceX is rethinking/designing their BFR ITS Mars vehicle. Musk may release in a few weeks.
    Will SpaceX consider Zubrins suggestions, will this make ITS a more competive System for hauling payloads to the moon?
    Don’t we need a tug from low Earth orbit to high earth orbit to build a cheaper highway to the moon and Mars?
    Seems NASA needs to scrap plans for to use SLS and start designing systems to help SpaceX go to Mars and Blue Origins New Glen Go to the moon.

    Don’t we need a sustainable highway to these two places??

    https://twitter.com/dtarsgeorge/status/852696001680990208

    Inner Solar System Highway

  • Plans for Mars (DRA 5) are for something like 6 +/- SLS launches per mission. ITS is too notional at present to build a national space policy around. So, is there a credible alternative for the US to go to Mars using non-SLS launchers? We criticize the SLS because of cost but what is the alternative with sufficient capability that the policy makers can plausibly support during this political cycle?

  • Kapitalist

    A higher flight rate might be difficult because it’s built on the Shuttle industry which ever only produced 5 pieces. They went for reusability instead of mass production. And now the reusability is scrapped, so… it sounds like a great compromise, doesn’t it.

  • Kapitalist

    The alternative to $23,000,000,000 on the SLS/Orion to date is 170 Falcon9/Dragon launches, 25% more than all shuttle flights, at $133,000,000 each. Or at half that cost for uncrewed Falcon9, 7,000 ton in LEO. That’s the mass of 15 ISS space stations.

    I think SLS will fly a couple of times , to save face for the corruptocrats involved in this huge sabotage against human spaceflight. But I hope that at least the Orion never does.

  • Kapitalist

    The cost estimates vary depending on who you ask, and how it is defined. Some say $43 billion, others $23 billion. One could diminish the lowest estimate by another factor of two, take half of that and subtract 50%, and IT STILL DOESN’T HELP!!!

  • Yes, but from a logistics perspective, does it make sense to launch something like 11 FHs for each Mars mission even if it costs less? I believe that there are ways to reduce the number of FH launches such as by reducing mass (e.g. smaller crew, recycle, aerobrake, ISRU, extended stay, etc). But who has come up with the architecture that does the math?

  • JamesG

    The “shuttle industry” produced lots of ETs and the SRBs and even the Orbiters themselves were pretty much completely rebuilt between missions. That was the “shuttle industry” SLS was intended to save.

  • pathfinder_01

    Depends on how long it takes to launch 6 SLS missions. At about 1.5 a year it would take 9 years for each mission! Even at 1 a year it would take 6.FH with reuse could have a faster pace than that. If FH launched as often as F9 did between 2014 and 2015 they would have the 11 launches needed.

  • Paul451

    but from a logistics perspective, does it make sense to launch something like 11 FHs for each Mars mission even if it costs less?

    The cost of a full FH launch isn’t listed, but the price of a 1/3rd capacity launch in reusable-mode is about $90m. Multiplying by 3 is $240m for a >50t expendable launch. It would probably be less, but there’ll be a “NASA tax” which puts us back up there, so let’s call it $250m each all-up. If you use 11 FH’s per mission, that’s $2.75b. That’s slightly less than one year of SLS/Orion funding. (With reusable mission hardware, the number of launches per crew-change should drop. But $6b launch costs per decade seems reasonable.)

    That theoretically frees up around $24 billion over the next ten years to immediately start funding mission hardware. Whereas the SLS-based architecture won’t have funding freed up until around 2030, and needs to kill the ISS to do so. Ie, it won’t start to seriously develop Mars hardware for at least another decade.

    An FH-based architecture allows them to start now.

    (Of course, the stupid way to develop mission hardware is the way they developed ISS. And the way they developed the Shuttle. And the way they are developing SLS and Orion. And the way they intend to develop DRA5.0 hardware.

    But if you’re gonna go commercial, go all in. Parallel programs to develop multiple “commercial” architectures for the moon and Mars. Contrast the cost of COTS & CCdev with the development cost of Orion just to date; and the apparent development cost of FH with the dev cost of SLS so far.

    Picture similar savings in every other aspect of the mission: Long duration deep-space ships, habitats, landers, resupply missions, etc.

    Deep-space ships are essentially space-stations with engines, habitats are space-stations on the ground, making the precursor “stepping stones” obvious. That possibly lets you retire ISS early, transferring ISS’s funding to doing science on multiple specialised commercially-run space-stations in LEO and lunar-orbit. (So cancelling SLS could produce multiple ISS replacements, instead of needing to cancel ISS just to make SLS viable.)

    Precursors for manned landers is, obviously, unmanned landers of various sizes. Years before the first manned Mars landing, you’d have multiple vendors able to soft-land 5-15 tonnes of hardware on Mars. Some might be capable of lunar landing as well. (SpaceX estimates that Red Dragon could land 2 tonnes of payload on Mars. MSL masses less than a tonne. Imagine what you could do with 5 tonnes, or 10.)

    $56 billion over the next two decades would buy a lot of COTS-like programs, all in parallel. All starting next year.)

  • Kapitalist

    How many launches has it taken to assemble the ISS? Even SLS needs assembly in space to land astronauts to the Moon. Like Saturn V was too small to go directly surface-to-surface, it had to undock and turn and dock with the lander en route, then undock again, land, ascent, dock, undock. Making multiple Falcon Heavy launches and assembly in LEO could require *less* assembly and be easier and safer than using one SLS launch.

    One shouldn’t launch crew on the SLS anyway. It has way too infrequent launches to ever get human rated by any ordinary standard. Better use a small commercial launcher with 50 or so straight uncrewed successes in a row. Also, the Orion is more massive without any benefits for it, and the SLS requires a 3 ton launch escape tower with crew.

    If there were a 100 ton piece of payload that could not be disassembled, then the SLS would win. But we don’t have that.

  • I don’t think that the ISS is the correct illustration. I would hope that we wouldn’t propose using many launches and in-space assembly like the ISS for missions to Mars. I consider docking components as different than “assembly” although I do recognize that it is a matter of degree. At some point one needs to be reasonable. Docking a few times is reasonable. Docking or assembling many times per mission is unreasonable. Where one draws that line is the fair question that I am posing.

    With more modules comes a a greater surface-to-volume mass, some operations cost, perhaps greater on-orbit propellant boil-off or greater need to prevent that, greater likelihood that at least one module will be lost per mission at launch, etc.

    I’m suggesting that those of us who advocate the more cost-effective use of Falcon Heavy for Mars surface missions need to put forward an architecture that shows how that can be done within reasonable limits.

    Does the lack of commenter agreement with this position mean that y’all don’t believe that we need to put a forward FH-based architecture for Mars — that it is so obviously true to all that no such architecture need be specified?

  • redneck

    I would argue that a Mars mission lends itself particularly well to multiple flights. The only requirement for large volume is the crew hab for the long legs of the trip. The vast majority of the launches will be propellant, landers, reentry gear, provisions, solar arrays, and sundry other articles that have no reason to need simultaneous arrival on a large vehicle.

    Even the large hab volume could possibly be done on a far smaller vehicle if upper stage tanks were specifically designed as wet launch, dry hab volumes. This could be one of the reasons for current interest in methane as it would be considerably easier to detox for living space than kerosene. Center core of a FH running on methane could reach LEO with minimal payload except for that 200+ feet long tank/hab.

  • pathfinder_01

    ISS on the american side used the shuttle which had it’s advantages(i.e. wider corridors due to not needing to put engines and other stuff needed for docking) and disadvantages such as everyting being dependant on a single rocket system.

    All architectures depend on the technology being available and it is really a matter of just how small an package does each part need to be(or can be made to be) and where you want to stage and where you want to depart. The problem with SLS is that politics is driving the train beyond the point where funding and technolgy can reach a reasonable solution. I mean you want to build a rocket today(2017) for a mars mission in 2030! How could you even define what properties you will need in the rocket for a mission that is 23 years in the future? Apollo only needed to plan 8 years into the future. At the moment the DSG is likewise in no state to be ready quickly enough to provide payloads to SLS.

  • Kapitalist

    Even Elon Musk seems to be waiting for the super-heavy launcher before sending humans to Mars. That would be a good answer to what the launcher size limit is for assembling a mission to Mars: It needs to be bigger than FH. According to Musk obviously bigger than SLS too.

    I think that today’s launchers are too small for putting together a crewed Moon mission. I think 4 launches assembled would be needed just to pre-place the ascent vehicle, 11 in total minimum. The many launches would cost almost as much as one SLS launch anyway. But the Falcon Heavy would be enough for a good two week human mission to the Moon. Two FH with the ascent vehicle and habitat separately pre-placed on the Moon, a third FH with the transfer vehicle and landing stage with which the crew in an F9/Dragon docks in LEO, should be enough. One single docking in LEO. Price tag for launches: $400 million.

    If one of the robotic missions fail, the material lost is smaller than for an all-in-one launch, and since there are daily launch windows to the Moon, there’s no extra delay beyond the replacement work.

  • Kapitalist

    Only 13 years until 2030. Time flies! 2033 is a good conjunction with Mars, meaning more months on the surface and fewer on the road, so say 16 years.

  • There’s nothing inherently good about more launches rather than fewer. In this case it’s really a question of whether the cost savings of using a different, smaller, but cheaper launcher (i.e. the FH) compared to the benefit of accomplishing the same thing with fewer launches. It’s a cost versus risk thing. If we can accomplish the same mission using a lower cost (kg to LEO) but smaller launcher and we can do it within reasonable risk then, yes, that’s the way that we should go. But I continue to ask, “Where is that architecture” or “Who can develop such an architecture”?

  • redneck

    Less risk with distributed launch on proven vehicles. No in space assembly required as there are no parts that require heavy lift, just docking. Cheaper is obvious. Schedule is much sooner.

  • Mr Snarky Answer

    I know why they are behind schedule, they are too distracted by side projects like reusability and going to Mars…oh wait.

  • By itself, distributed launch is more risky and expensive. Here’s an illustration. Say one wishes to put a 60 ton satellite in LEO and one has two options:
    1) One Falcon Heavy (64 tons) or
    2) Three Falcon 9s (22.8 tons X 3).

    A single FH costs less to launch than launching three F9s, correct? A 60 ton satellite composed of three 20 ton modules requires two in-space dockings whereas a single 60 ton satellite launched on the FH requires no dockings. Two dockings carry more risk than no dockings, correct? Three modules requires two docking adapters (more mass) and more maneuvering propellant. Failure to place any of the three modules in exactly the right orbit could make that module useless requiring an additional launch of a module replacement. Three launches require more launch range fees, more launch crew time, etc. Whereas a greater launch frequency lowers the per-launch cost launch, it increases the overall cost.

    As I said previously, by itself, there is nothing inherently good about an architecture requiring more launches than fewer. Instead, the actual comparison that we are dealing with is between a low-cost, commercial launcher (i.e. Falcon Heavy) and a very expensive big government launcher (i.e. SLS). It’s that difference where the problem lies.

    So my question stands. Which is better? An architecture using a more expensive launcher which requires fewer launches or an architecture that uses more launches but at an overall lower cost? I favor the latter but until someone publishes such an architecture, I don’t think that we’ve made our case.

  • redneck

    It is not yet proven that the FH will be reliable or cheaper than F9, likely yes, proven no.

    There are no 60 ton satellites at this time. There are not likely to be any until affordable and reliable launch is available.

    We seem to be in agreement that planning anything around SLS is stupid on steroids.

  • ReSpaceAge

    🙂

  • Paul451

    Belatedly,

    With more modules comes a a greater surface-to-volume mass, some operations cost, perhaps greater on-orbit propellant boil-off or greater need to prevent that, greater likelihood that at least one module will be lost per mission at launch, etc.

    It’s worth emphasising that none of the modules in either the SLS lunar orbit architecture, nor in DRA5.0 Mars architecture exceed 25 tonnes. Ie, none are too heavy for FH in a partial-recovery mode.

    Hence none of the objections you raise apply. For rough comparisons, there’s no difference between architectures using SLS and FH, only the cost of the launches.

    (In reality, the one big exception is the wide-diameter heat-shield assumed for the Mars landers. However, I can’t see that working around this single item is worth the extreme cost of SLS/Orion. For eg: If it, like Orion, is assumed to use tiles, a three part folding heat-shield can be launched to ISS and deployed and gap-tested before vehicle assembly. And obviously NASA is already working on an inflatable heat-shield. Or a longer-than-wide biconic lander. Or…

    A smaller exception is that it is probably more optimal to add a long-duration habitat module to Dragon than developing a long-duration version of Dragon. (Dragon will already have long-duration stand-by capability for ISS ops.) But again, the cost of such a hab-module is massively eclipsed by the cost of Orion. Plus once you have such a hab-module, it will give you a more flexible architecture. A 21day version of the module (for lunar flights) gives you a stepping stone to developing a Mars version (which is necessary for Orion-to-Mars anyway). As a bonus, the hab-module lends itself to a COTS/CCdev funding model.)

  • Paul451

    Belatedly,

    Even Elon Musk seems to be waiting for the super-heavy launcher before sending humans to Mars. That would be a good answer to what the launcher size limit is for assembling a mission to Mars: It needs to be bigger than FH. According to Musk obviously bigger than SLS too.

    Musk wants to build a colony ship. 100 people per flight.

    That’s not comparable with architecture optimised around a 6-man science mission.

  • Paul. In general I agree with you and want to see a FH-based architecture for Mars. Can FH place 25 tonnes modules at a cis-lunar point? Can U refer me to that info? Once again, how many FHs would it take to go to the surface of Mars and where has that architecture been written up?

  • Paul451

    Can FH place 25 tonnes modules at a cis-lunar point?

    That’s the point, they wouldn’t have to. To get a 25 tonne module into cis-lunar space requires two elements: a launcher to get the 25t-module into space, and a way to get that 25t module out to cis-lunar space. But the two elements don’t have to be on the same vehicle.

    And, by an amazing coincidence, the DRA5.0 architecture requires a stack of modular booster-stages. It is a necessary development for SLS-based Mars architecture, precisely because SLS isn’t capable of lofting a Mars ship in a single- or double-launch. SLS requires modular design and in-space assembly!

    Such a booster-stage could therefore be used to ferry the other modules to the final assembly point. And given that additional job, it lets you properly test those modules in a way that the SLS-based architecture probably can’t afford. (And you would find several vendors itching to bid on such a beastie starting immediately, if the funding were available.. LM is itching for someone to pay them to develop ACES, for example. Under SLS/Orion, the extra funding only possibly becomes available after ISS is splashed.)

    how many FHs would it take to go to the surface of Mars

    One.

    But my point was, the DRA5.0 Mars architecture is already modular, allowing you to directly contrast FH and SLS.

    FH might require double the number of launches for the same mission, but the cost is so very much less, you don’t care. If it was triple the number, it wouldn’t matter. Two years on SLS development funding and your launch costs are paid for. Leaving the next 20+ years of development funding available for the actual mission hardware.

    It may turn out that an optimal FH-based architecture would be different, but you can simply duplicate the existing SLS-based model. And doing so shows that there’s no irreplaceable benefit from SLS, even in an SLS-optimised mission architecture.

    Likewise, you don’t have to depend on FH. New Glenn and Vulcan both loft around the same amount. A COTS-style fixed-price fixed-delivery program would be possible to add redundancy and launch frequency.