An Electric and Steel-making Way to Build a Science Station on Mars

[RifMilesOlsen]

The Journal of the British Interplanetary Society (JBIS), which published Zubrin and Baker's Mars Direct in 1990, published
a new plan for Mars a few days ago (on September 7^th, 2024). It is called

The Ironberry Plan: an Electric and Steel-making way to Build a Science Station on Mars.

The Ironberry Plan is flexible enough to include many ideas and activities from others. However, it has some characteristic early steps that
make this plan distinct. Those characteristic steps are: 

1. The robotic harvesting and processing of a superb iron ore at the iron region of the Meridiani Planum (IRoM);
2. The use of Mars-optimized, robot-controlled steel-making and manufacture of steel parts;
3. At least for some years, a deliberate push for mutually feeding exponential growth in electricity generation, steel-making
and steel manufacture, where early steel manufacture is focused on making more electrical generation equipment;
4. Using the rapidly growing power capacity and advanced robotics and processing of local sediments, the start of many
more foundational activities (i.e., water liberation, sulfur-concrete construction, multi-material manufacture, agriculture and
more);
5. Using 1-4, the building of a science station at IRoM while also converting excavated sediment caverns into underground,
radiation-protected spaces for the science station.

I am the paper's author. I look forward to reading reactions to these characteristic steps in the forum. The 6000+ word paper should cover
many of your possible points, comments and questions. Although, there is so much to cover in a comprehensive
plan to build a science station at IRoM, that the paper is only a outline. It should be easy for many people to find and discuss numerous details
not covered in the paper. I am currently funded by the National Science Foundation to develop methods and reactors for steel-making at IRoM,
 so feel free to ask me tough questions on how this will be done. 

You can find the paper at

https://doi.org/10.59332/jbis-077-03-0091
(Free abstract read, 5 pound sterling download )

and
https://www.researchgate.net/publication/383857204_The_Ironberry_Plan_an_Electric_and_Steel-making_Way_to_Build_a_Science_Station_on_Mars
(Free download,  please tell me (https://twoplanet.life) about a free download so I can pay JBIS)
 

In overview, the Ironberry Plant features intense ISRU in a favored location, robotics, and rocket transformation. It needs large rockets, but it does not require large numbers of large rockets, or rapid rocket reusability (RRR), or RRR's need for large consumption of precious Mars water. It offers an alternative to NASA's Moon-to-Mars policy in that, at least for an extended period, its activities do not focus on transporting humans to Mars or the moon. In contrast, it uses very low cost precursor robotics, manufacturing and local resources to provide an impressive infrastructure (for example, a robotically-built landing pad) that can support humans on Mars (possibly hastening our arrival there) and also continuations of Carl Sagan's Water Strategy, and the beginning of astronomy from Mars (hopefully in the middle infrared, which is best for imaging watery exoplanets).       

[MaddieMac]

I thoroughly enjoyed reading this paper; it offers a clear overview of a compelling and seemingly feasible plan for future development on Mars. While I'd previously heard of the availability of iron ore on the planet, it was interesting to see how this fact, combined with steel-making, robotics, and electricity have the potential to lead to significant achievements on Mars.

[Twark_Main]

There's a lot of fantastic research and detail in this paper, thanks. Anyone with a serious interest in colonization is missing out if they don't pay heed! 


That said, I'm not totally onboard with the self-reinforcing loop part of the premise.

Electricity helps make steel? That half of the "flywheel" makes sense to me.

Steel helps make electricity? That's where I foresee problems. High-concentration solar seems to be a dead-end technology on Earth, and all the examples I can find use fresnel lenses mounted on the panels. They just look like a regular (if large) tracking PV panel, not the Solar Power Tower type design. And again, even that technology hit a dead-end in the mid 2010s.

Concentration doesn't work great on Mars either, because you can only use the beam (not diffuse) component of insolation.

It's easy to roll out ultra-thin panels which self-clean any dust via an electrostatic grid. These ultra-thin panels require very little mass delivered from Earth (certainly less than existing PV concentrator designs), and it doesn't require a bunch of Martian steel (which isn't free either).

If the other half of the self-reinforcing cycle (steel helping to make electricity) doesn't really work, or maybe is non-zero but isn't a very strong effect, then the bold promises enabled by an exponential growth loop come crashing down.   



I love the paper, and I do think the thesis can be considerably strengthened with only a bit of re-work. There are self-reinforcing loops (the meat of the paper), but it's more that steel enables...   machinery to mine & process more steel.  Admittedly this is not as cool as the PV -> steel -> PV story, but I think it's more technically feasible and academically defensible.

[RifMilesOlsen]

Hi Twark_Main

Thanks for challenging a basic premise of the plan. There are straightforward responses to your points.

First, to an extent, you are right. On Earth, the economic solar technology has turned out to be flat photovoltaic panels
mounted on fixed support (i.e. that do not track the sun's direction). On Mars, this will not be the best solution, at least not
for some years. The economically optimal  solar technology on Earth depends on a large, sophisticated solar industry.
In other words, these Earth-optimal solar panels are not easy to make. It will take a considerable build up in manufacturing
expertise on Mars before they will made made and deployed on Mars. My gut feeling is that the flat panel, no-concentrating
solar panels will ultimately be the preferred solar technology on Mars. Further, your suggested self-cleaning with an electrostatic grid
is an excellent feature to include.

     
You are on somewhat shaky ground  with your doubts about the feasibility of concentrating solar technology. There are various flavors
of concentrating solar technology. All but one of these flavors has either not been seriously demonstrated or they have been built
at utility scales but have proven more expensive than economically optimal PV. (Side note, the increases in PV panels and lowering of
manufacturing and installation costs of these panels has been absolutely fantastically good, which proved hard-to-compete with for concentrating solar, and shout-it-from-the-rooftops good for everyone on Earth.). I said all but one flavor. These is one form of concentrating solar power
that is thriving in Australia.  It uses a "power-tower" concentractor architecture, a cooled, high intensity PV panels (that can handle >1000x
of normal insolation on Earth) for a solar receiver, and also reuse of heat generated at the solar receiver. This solar technology is
built by Raygen. You can see it in action at three scale's on their website. I like the smaller (250 kW) scale to start off on Mars,
which was built (times 3) for a first pilot plant (https://raygen.com/newbridge/) and cited twice in the paper. Raygen have now built
a larger 4MW plant (https://raygen.com/carwarp/) and have a 150 MW plant in development (https://raygen.com/yadnarie/).
Other not-so-good power tower plants were successfully built and operated at Ivanpah in California and in southern Spain.
Please look at the 29 second video at   https://raygen.com/newbridge/ to see the three 250 kW units in action - These 250Kw
units are models for what might be constructed robotically with locally made steel or with steel recycled from the propellant tanks on a
transformer FirGStoL (which is probably a customized SpaceX Starship) - you should read the last section of the paper.

In the Ironberry Plan the concentrating PV reciever units (that can take > 2000x Martian insolation levels)  and various control units
and hard-to-make parts are transported from Earth to Mars. While the very large majority of the mass of a model 250 kW unit is made
of simple parts manufactured locally by robots using steel, firstly recycled from the FirGStoL, afterwards locally made steel.

I should point out another advantage of the Raygen solar concentrator model, it produces very large quantities of useful process
heat. This process heat will be extremely useful for liberating water from hydrated rocks (see my paper  https//doi.org//10.1061/9780784484470.032) and driving some of the processes to turn the local IRoM sediments into useful derivatives (see the IronBerry
Plan paper for an introduction to this.)

Remembering that we will not be able to manufacture solar PV panels on Mars anytime soon, the Raygen model establishes the feasibility of using concentrated solar photovolatics at IRoM, while the local resources, Mars-optimized steel-making, and robotics should make
manufacturing many Raygen units a practical task, and hence the Ironberry Plan survives Twark_Main's challenge.

   
 

   

[Twark_Main]

First, I should say I'm genuinely trying to "close" the engineering. I'm dismayed because I'm failing to make it work, not trying to tear it down.   

I agree with your point about PV manufacturing on Mars being a later technology, but it seems like thin-film "rolls" of solar panels delivered from Earth (and later possibly laminated on Mars with in-situ plastic to reduce imported mass further) will still require less upmass from Earth. Modern thin-films are incredibly thin and light, and even just the motors and power electronics required for pointing the same area of mirrors would require more mass.  Ouch. So my problem is I can't make it work even assuming all the Martian steel manufacturing and shaping and assembly is free, which seems a tad optimistic... 


I'm not saying that Martian PV beats imported concentrating PV kits, I'm saying that imported roll-out-and-stake-down PV beats imported CPV kits.


Mars optical depth (tau) runs about 0.6-1.0, so even outside of dust storm season you're losing about 1 - e^-0.6 = 45-63% of incoming sunlight. This is after losing ~half because Mars is further from the Sun. During even a mild dust storm the production will rapidly drop to zero, because concentrating PV can only use beam sunlight not diffuse sunlight. Roll-out panels don't have this disadvantage: they can use diffuse sunlight. A system sized for ISRU propellant production can still produce a few percent power during even a severe dust storm, enough to run minimal life support and essential services. This reduces strain on batteries and stored supplies.

Not a show-stopper by any means, but it's just yet another multiplier I find myself fighting against to make the system close.

[RifMilesOlsen]

Hi Twark_Main,

First, by a scrolling problem, I did not your original compliments - many thanks for those .
It is great to be having a conversation.

Your points are centered on improving the ideas behind the Ironberry Plan. That is good, I like improvements.
 However, I want to point out to other readers of this post, that by implication you have admitted that building Raygen-type
concentrating solar electricity plants using a lot of locally-made steel could implement a electric/steel loop
and that when such a loop is implemented it can be run to produce exponential growth in electricity and steel production.         

You are focused on improving electricity production, and you see thin-film rolled PV sheets as a superior technology for improving
Martian electricity production. You make a solid point about PV sheets being able to work in diffuse light as well as with beamed sunlight.
(This advantage of (unconcentrated) PV over concentrated solar is a principal reason why I think we may revert to unconcentrated
PV in the long term, i.e., after we can set up a factory to make PV's on Mars.) I also think the very first solar unit (~3 Kw) that should
be set up at a FirGStoL should be a fold-out (unconcentrated) PV unit that powers very early robotic movements.  You also point out
the extreme low mass to solar production ratio of thin-filmed PV sheets, another excellent point. However, there is still a long way to
go to that show building electricity generation plants on Mars using transported thin-filmed PV sheets fundamentally undercuts
the basic Ironberry Plan loop where complimentary electricity and steel production leads to exponential growth in both.  This loop
may possibly be improved by replacing a Raygen-type concentrating solar power units with solar generation units made using
thin-filmed PV  sheets transported from Earth that are mounted on supports made from locally made steel (these sheets will need
supports and they would likely be made of steel, see my note about plastics below). The Ironberry Plan
might also be modified to start by building both types of generation plants and then learning which of the two types is best for
optimizing our abilities with exponential growth. 

I will add a word of caution, locally-made plastics will be scarce. You need sources of hydrogen to make plastics. Water is the only
known Martian source of hydrogen of any size, and this water is itself very scarce.       

[deadman1204]

Alot of the "automated robots running factories" part of plans always gives me pause.
We cannot automate mining or such here on earth, where it would be incredibly easier compared to doing so on mars. And were not "close but not quite", its not even on the horizon.

It feels very hand wavey to say "we'll just have automated robots do it'. I hate to call it mars fan fiction with lots of details, but it assumes an awful lot of technology and waves everything away as easy to solve "engineering problems". This imagines robotics that can function under mars conditions that are generations ahead of the best we can do on earth. Even if we could do this on earth, it would be several gigantic leaps in technology to replicate the same thing on mars with people there. Adding full automation without any human intervention? It gets even worse.

[RifMilesOlsen]

Hi deadman1204,

Thanks for reading and replying with a comment.

You are right to skeptical of "automated robots running factories." For example,
in the past several years Tesla attempted a fully robotic car assembly line. They failed (at least a few years ago)
to replace most of the human autoworkers with robots.   

A response comes from three directions:

1) the Ironberry Plan does NOT need whole systems to be run fully automatically with robots (and the
paper does not suggest this).  However, to avoid chronic problems with communication latency with Earth, it  does
need some significant levels of semi-automony from the robots and the ability to communicate long series of
commands to robots in a short-hand, very high-level code. 
2) There will be extensive supervision and guidance of robot action by engineers on Earth.   
2) There are some key Ironberry Plan robotic tasks that can be run with a high degree of autonomy. Such tasks
are relatively simple, repeated many times and well-known from trials of these tasks in test workshops here on Earth.

In the third direction, all tasks for the normal operation of an iron- and steel-making factory will of this repetitive kind that
can thoroughly worked out and tested here (on Earth). Note, such a factory will be enclosed in a structure (probably in a transformed
spacecraft) where the operating robots will either be stationary or mobile on rails or flat surfaces.
Granted something will go wrong with implemented such tasks. But reliably sensing when something has gone wrong
is entirely possible and the first and second direction capabilities will be enough to fix almost all steel-making problems
when they occur.
   
In the first direction, our increasing expertise with AI will provide impressive levels of semi-autonomy;
for example, a roving robot will be able move and navigate itself quickly across longish distances (say 100m-1000m)
of the favored "smooth sheet" terrain at IRoM/Berry Campus (such terrain is remarkably flat). On high-level coding,
we will need robots capable of carrying out a series of commands coded at such a high level that the code
comes close to natural human language; for example, "goto the outer viewing ring of heliostat #41, and take a
360 degree video of the state of this heliostat, send the video to Earth base, assess whether it is possible to goto the
inner viewing ring of heliostat #41, if yes, then goto this inner viewing ring, take a 360 degree video of heliostat #41 from its
inner ring, then send this video to Earth base." Such high-level coding, and robotic implementation of commands given in such a code,
is surely within our AI assisted reach today, and especially so when the environments the robots operate in (such as inside a
field of heliostats) is accurately mapped.