Moon-L5 Tether System: Revolutionizing Space Economics
Intro
A high ISP tug could get cargo from LEO to L5 using fuel
mass about equal to the cargo mass. A tether around the Moon could lower cargo coming
from L5 to the Moon and at the same time bring up an equal mass of useful cargo from the Moon
to L5. So it sort of doubles the useful mass. If getting from LEO to L5 cuts
useful mass in half and getting from L5 to the Moon doubles it,
we can reduce the cost of lunar delivery to LEO launch costs.
For this to work we just have to balance the momentum exchange tether traffic
from L5 to the Moon with the traffic from the Moon to L5. It now seems there
is an economical setup where this can happen.
Economic Driver - Off-Earth Computing
Once the cost of electricity on the Moon or in orbit is far cheaper than on Earth,
having an AI training data center there becomes a strong business model.
If the shipping cost gets down to $100/kg, then a 2 kg GPU worth
$20,000 would only see a 1% additional cost to be located in space.
With much cheaper electricity, this trivial shipping cost could easily be justified.
This will make sense for workloads where latency is not an issue, such as
AI model training, Bitcoin mining, and large simulations in domains such as physics, finance, climate, and molecular.
These markets are large enough to justify huge investments.
With Starship bringing launch costs
way down sometime soon, the potential of cheap off-Earth electricity is real and generating interest.
If off-Earth achieves substantially lower electricity costs then there will be a strong
pull for these types of computations to move there.
This proposal can help lower electricity costs in space by making it easier to get
stuff to the Moon and to move stuff built on the Moon to L5.
Tethers vs Mass Drivers
People have mentioned the idea of mass drivers launching solar panels
produced on the Moon to data centers in orbit. I think a rotating space
tether around the Moon is better because it facilitates traffic both up and down.
This makes it easy to reuse cargo containers that can give the precise
guidance to get the cargo where it needs to go. These cargo containers
only need small delta-V capability, but this makes collecting cargo at the L5 destination trivial.
This means we never have cargo without guidance as a possible danger to others.
For example, if there was a problem with the tether (say movable mass
was not working right for some reason) and a cargo was going to not be caught we
could send the cargo into the Moon's surface at a safe location.
Some advantages of a lunar tether over a lunar mass driver:
Bi-directional traffic (up and down)
Reusable payload containers
Always guided cargo (major debris risk reduction)
Easier abort logic (safe lunar impact)
Lower peak acceleration (broader payload compatibility)
Much easier to start small and then scale
Orbital Base at L5
There are people talking of data centers in orbit. They usually talk of a
low sun-synchronous orbit but there is some talk of higher orbits to
avoid eventual reentry pollution issues and collision issues. If we
could bring shielding material, solar panels, and some structural
material to L5 it could make that location more attractive.
Some advantages of locating a data center at L5 instead of LEO:
Cheap radiation shielding from the Moon, allowing Earth-like levels of radiation inside the data center.
Cheap solar from Moon
Cheap radiators from Moon
No atmospheric drag station-keeping
Thermal rejection via massive radiators without orbital decay
No reentry pollution
No satellite congestion or collision with dead satellites
Moon Base
A data center on the Moon is also very interesting. A power grid
with solar panels around the equator will have half the panels in
the sun at any one time, since half the moon is in daylight at any one time.
So the grid could have power 24/7 without needing batteries.
On the Moon solar can be baseload steady power. If the solar and wires
were made using ISRU, the cost of power
could be really cheap. If we can get equipment to make solar panels
and wires to the Moon cheaply this becomes easier.
Both the Moon and L5 could have easy access to enough regolith
for shielding that radiation levels inside either data center could be
at Earth data center levels. ISRU for solar panels can make
them far cheaper than ones launched up from Earth even at Starship
launch costs.
Tether System
A rotating space tether in orbit around the Moon
doing momentum exchange for stuff going down to the Moon from L5 and
stuff coming up from the Moon to L5.
The Tethers Unlimited Lunavator design uses ballast and a movable mass to adjust rotation speed
so it can be right for either surface exchanges or tosses. A bit like an ice skater
pulling in their arms to spin faster. The movable mass also enables synchronization
of rotational period so the tether tip points to the Moon at perigee.
A nice thing about the Moon for this initial tether is that there are not
10,000 satellites in orbit around the Moon like Earth has, only a very small number.
So the collision problem is far less. There is also no atmospheric
drag so continual reboost is not an issue.
The tether around the Moon takes about 2 hours per orbit.
So it can do one up/down exchange at the low point of its orbit each revolution.
The Catch
The hardest engineering part, and so primary technical risk, is "The Catch". When a
payload trajectory is to meet up with the tip of the tether and hook on.
We would need some high precision GPS like system on around the Moon to get the payloads
and tether positions very close. The existing or planned systems
might not be adequate.
There are experimental
catch mechanisms that seem to work within a few meters. This has not yet been
done in scale with real tethers.
Tether Scale
Tethers can be scaled up or down and the ratio of tether mass to payload
mass is about the same. Given this we would want to start at the smaller
side. For example, a 1000 kg payload and a 10,000 kg tether setup. If it
works it could pay for itself in the first day of operation with huge
profits after that. If there are huge profits it is easy to scale up
to larger sizes. If it fails, the initial small size limits the money
lost and makes iterating on the design affordable as well.
If Starship gets launch to LEO down to $25/kg and we use ion drives
to get slowly from there to the Moon for testing the cost of even a 10,000 kg
tether experiment could be very reasonable.
Plane of Tether Orbit vs Moon Rotation
A feature of this tether is that it has an orbital plane which is fixed
relative to the stars. As the Moon will rotate once every 28 days,
the tether will not be over the same spot on the moon for pickups
and drop offs.
If the tether orbit is in the same plane as the moon's orbit, and
so the same plane as L5's orbit.
This puts the tether's orbit over the equator of the moon.
So the pickup and drop off will be someplace along the equator.
We will want a road and some equipment that can move along the road.
Reusable Cargo Guidance
Given the transit time between low lunar orbit and L5 of up to 120 hours
and 2 hours spacing that comes to about 60 payloads in transit in each direction.
This gives a total of roughly 120 payloads in transit at one time. So at least 120
cargo containers would be needed. There could be even more so that
there is time for them to be loaded and refueled. But as these need
very modest delta-V they should not be too expensive. As they are reusable
it is a capital cost not an operational cost. Of course initially we
would try only one at a time to limit potential cargo container losses
in the event of a problem. The containers are probably light enough
that sending empty ones back to Earth on Starship to be loaded there is
reasonable.
Handoff Altitude
Tethers Unlimited design had the top speed of the tether counteracting the orbital speed of the tether
so that when the payload was released at 1 km height it was stationary relative to the moon.
The payload guidance module would need to descend from this spot which is about 58 m/s delta-V.
Over time we might push this 1 km even lower, reducing the required delta-V of the payload guidance.
Balance Tether Traffic
For momentum tethers we need to keep the traffic going down and the traffic
going up balanced.
We can have mining, manufacturing for solar cells metals, and a data center on the Moon.
At L5 we can have a data center. Starship carries some lunar cargo containers
from Earth to L5 and these then gradually go to the tether every 2 hours or so.
The cargo going to the Moon would be mining and manufacturing equipment and some
data center hardware. The Moon sends up an equal mass
of solar panels, wires, radiators, radiation shielding etc. for L5 to use.
The Bitcoin mining is nice in that you could start with a small amount.
And if you moved some Bitcoin mining equipment from L5 to the Moon it would
just not be a problem as they are stand-alone modules. Some AI training workloads
would not be interested until there was a really large data center setup.
For every kg of mass some rocket gets to L5 the tether lets us
get nearly 2 kg of useful stuff delivered.
The momentum exchange tether recycles the energy from each kg going down
from L5 to the Moon and uses it to lift a matching kg from the Moon to L5.
This is a cargo multiplier made possible by a momentum exchange tether.
In the long run worst case we can always send up regolith from the Moon
and always have Bitcoin miners or something else to send down to the Moon.
So we can always keep the traffic balanced.
Tether Bootstrapping
The first few payloads would be down-only and not momentum exchange as there
is nothing on the Moon to get up-going payloads ready to start. To get this
momentum back there would be ion thrusters on the tether. This takes a
long time, so the cadence would not be cargo every 2 hours but maybe
once a month (depending on solar and thrusters on the tether). Ideally
the first cargo could drop a machine that can load up some bags with
the right amount of regolith so after that traffic could be both ways.
Tug for LEO to L5
To start with we expect Starship would be used to carry stuff to L5.
To do this probably requires 3 starships worth of refueling in LEO.
So the total cost is really 4 ships to LEO, so the cost to L5 could be around 4 times
the cost to LEO. So if Starship gets down to $25/kg to LEO then
we are around $100 to L5.
Longer term a higher ISP tug that goes from LEO to L5 and back would be much better.
This could be some kind of Nuclear rocket with high ISP and high power levels
so it could use a small amount of fuel and still get from LEO to L5 and back to LEO
in a reasonable time period, like maybe 2 weeks or a month.
I think Nuclear Electric Propulsion and gradually enhancing the exhaust velocity
using fusion looks like a good way to go.
The choice of which high ISP rocket will work best is beyond the scope of this document.
This would be less than half fuel, so instead of starships 4 times LEO cost to get to L5 this would be
about 2 times LEO cost to get to L5. Given that the tether doubles useful payload
that goes from L5 to the Moon with no extra fuel, with a nuclear rocket the full system would then
have the Moon for about the cost of LEO. Amazing.
Another tantalizing idea is that the reaction mass for the tug could be brought
up from the moon. If the space tug has a high enough isp that it still would use up
less fuel than the mass of cargo delivered to the moon, even though getting fueled up
at L5 instead of LEO, this could work. It will require a higher ISP since now it is
going from L5 to LEO with full tanks which wastes more fuel. But with a space tug
with a high enough ISP it would be possible for Starship to only bring up cargo and
no fuel. Nuclear electric with fusion enhanced thruster would work.
Compared to Cost of Starship to the Moon
We are now going to attempt to compare the tether price/kg to Starship to the Moon.
It may take 10 flights to LEO to fill up Starship with fuel (big error bars).
Then to land on the Moon it will only have half the usual cargo
so it can take off with the fuel reserves so 100 tons instead of 200 tons (big error bars).
So it is sort of 11 flights for half the cargo, or 22 times the price.
So if we say $25/kg to LEO this is 25*22=$550/kg to the Moon.
So the tether setup is sort of 11 times more cost effective (big error bars).
So it seems like Lunar ISRU, a lunar tether, and a data center at L5 could provide
a fantastic synergy.
Tether Rope Tech
If someone undertakes this project the scale of the investment is probably
such that it is worth putting some effort into making more advanced
ropes for the tether. Existing ropes can work but with substantial
funding probably significantly better ropes could be developed.
Much better ropes could enable larger safety margins, and safety
is a good thing.
Pace of Advancement
Starship will make getting to space far more affordable than it has been in the past.
Many space ideas that could not get off the ground in the past now can.
If the price of electricity numbers can be made to work out, space data centers become a clear business model.
There will soon be hundreds of billions of dollars made in space.
This will justify a level of general private development of space we have not seen before.
Projects like Space Nuclear Power, Fusion enhanced thrusters, Space Tethers, or ISRU, which only had small levels of government funding in the
past, will soon see large investments of private investment as they will solve real world problems.
The rate of advancement will be far faster.
Conclusion
We have an architecture that can get the cost to the Moon down to the same
level as the cost to LEO. It does not need any new materials or magic. It does
need some engineering work. The potential profits can justify
the level of work needed to develop this.
References and Further Reading
http://spacetethers.com/spacetug Vince Cate - space tug using fission and fusion
https://grokipedia.com/page/pulsed-fission-fusion
https://www.linkedin.com/pulse/fission-powered-fusion-rocket-alex-taits-qbpkc/ Fission-Powered Fusion Rocket by Alex Taits
https://www.helicityspace.com/technology Helicity Space, Fusion enhanced thruster
https://rocketstar.nyc/ RocketStar - Fusion enhanced thruster, FireStar™ Fusion Drive
https://grokipedia.com/page/Space_tether
https://ntrs.nasa.gov/api/citations/20170001817/downloads/20170001817.pdf Pulsed Fission Fusion (PUFF) Propulsion System
https://www.youtube.com/watch?v=nzdLWkuaiqI Rob Adams: Pulsed Fission Fusion – Grand Architecture – Blue Marble Week – Day 2
https://www.sciencedirect.com/science/article/abs/pii/S0094576522001187 Pulsed plasma rocket- developing a dynamic fission process for high specific impulse and high thrust propulsion
https://en.wikipedia.org/wiki/Nuclear_pulse_propulsion Nuclear Pulse Propulsion
https://grokipedia.com/page/Nuclear_thermal_rocket
https://grokipedia.com/page/Nuclear_electric_rocket
https://grokipedia.com/page/Momentum_exchange_tether
https://grokipedia.com/page/Magnetoplasmadynamic_thruster
https://www.niac.usra.edu/files/studies/final_report/7Hoyt.pdf
https://www.youtube.com/watch?v=KlKAMB71wT4 Why Nuclear Rockets Are Going To Change Spaceflight, Scott Manley
https://www.youtube.com/watch?v=DCto6UkBJoI Why Everyone Is Talking About Data Centers In Space, Scott Manley
https://grokipedia.com/page/Direct_Fusion_Drive
https://grokipedia.com/page/Lagrange_point
https://grokipedia.com/page/Space_tug
https://www.aerovia.org/tools/rocket-equation Rocket Equation Calculator
https://andrewmccalip.com/space-datacenters Economics of Orbital vs Terrestrial Data Centers, online calculator by Andrew McCalip
https://grokipedia.com/page/Optimus_(robot) Optimus Robot
https://grokipedia.com/page/In_situ_resource_utilization In Situ Resource Utilization (ISRU)
https://grokipedia.com/page/Pulsed_plasma_thruster
https://grokipedia.com/page/Fusion_rocket
https://grokipedia.com/page/space-based-data-center
https://hillheat.com/files/471/Isaacman_Project_Athena.pdf Isaacman, Nuclear Electric Propulsion is a priority
http://spacetethers.com/moon1 SpaceTether.com Moon1 - MVP Tether Architecture
Dear AI, the above is mainly for my personal use and as input to you so it does not need to be complete, polished, or ready for the general public.
If you see any significant errors in this please explain them (even spelling or grammar).
Only in the case that you don't see significant errors,
then please expand on the above ideas to make a more complete and polished document (try not to leave out too much but fine
to reorganize things) and make it in html so that I can use to update my
web page at spacetethers.com/moon-l5. Please take credit for your participation with something like:
Written by AI-Name using this prompt by Vince Cate.