This proposal outlines an economically viable architecture for dramatically reducing the cost of lunar delivery to approximately the same level as Low Earth Orbit (LEO) launch costs. The system combines three key technologies: high-ISP tugs for LEO-to-L5 transport, a momentum exchange tether system around the Moon, and In-Situ Resource Utilization (ISRU) on the lunar surface. When properly integrated, these technologies create a synergistic system that could reduce lunar delivery costs by roughly 11 times compared to direct Starship lunar missions.
The fundamental insight driving this proposal is elegantly simple. A high-ISP tug can transport cargo from LEO to the Earth-Moon L5 Lagrange point using propellant mass approximately equal to the cargo mass, effectively halving useful payload. However, a momentum exchange tether orbiting the Moon can simultaneously lower cargo from L5 to the lunar surface while lifting an equal mass from the Moon to L5, essentially doubling the useful mass transported. When these two effects are combined—halving on the way up, doubling on the way down—the net result approaches parity with LEO launch costs.
The key to making this work is balancing the tether traffic in both directions. Cargo must flow both from L5 to the Moon and from the Moon to L5 in equal masses to maintain the momentum exchange. As we'll explore, there are compelling economic reasons why such balanced traffic flows will naturally emerge.
The entire architecture depends on a strong economic driver to justify the substantial initial investment. That driver is off-Earth computing, specifically workloads that are energy-intensive but latency-tolerant.
Once the cost of electricity in space drops substantially below terrestrial rates, establishing AI training data centers, Bitcoin mining operations, and large-scale simulation facilities beyond Earth becomes economically compelling. Consider the economics: if Starship achieves its target of $100 per kilogram to orbit, a 2-kilogram GPU valued at $20,000 incurs only a 1% shipping premium to operate in space. With dramatically cheaper electricity costs, this marginal shipping expense becomes trivial.
Several computing workloads are particularly well-suited for off-Earth deployment. AI model training requires massive computational resources but tolerates latency measured in hours or days. Bitcoin mining similarly demands enormous energy but operates asynchronously. Large-scale simulations in physics, climate modeling, financial analysis, and molecular dynamics all share these characteristics. These markets collectively represent hundreds of billions of dollars annually, providing more than adequate justification for major infrastructure investments.
SpaceX's Starship is poised to reduce launch costs from thousands of dollars per kilogram to potentially as low as $25-$100 per kilogram. This order-of-magnitude cost reduction transforms the economics of space-based activities that were previously untenable. The potential for cheap off-Earth electricity is already generating serious interest and investment. If space-based operations can achieve substantially lower electricity costs than terrestrial facilities, there will be powerful economic incentives driving these computational workloads into space.
Previous proposals have suggested using mass drivers to launch lunar materials into orbit. However, a rotating space tether offers several compelling advantages that make it the superior choice for this application.
Advantages of lunar tethers over mass drivers:
Bi-directional traffic: Unlike mass drivers that can only launch material upward, tethers facilitate both ascent and descent, enabling true momentum exchange.
Reusable payload containers: Guided cargo containers can be reused indefinitely, providing precise navigation and control throughout their journey.
Always-guided cargo: Every payload maintains active guidance, dramatically reducing orbital debris risk. There are no uncontrolled projectiles that could endanger other spacecraft.
Easier abort logic: If a tether catch fails for any reason, the guided container can be safely directed to impact the lunar surface at a designated location.
Lower peak acceleration: Tether systems subject payloads to much gentler acceleration profiles than mass drivers, allowing a broader range of cargo types including delicate electronics.
Incremental scaling: Tether systems can start small and scale up gradually as demand grows and operational experience accumulates.
The system employs a rotating space tether in orbit around the Moon, performing momentum exchanges between payloads descending from L5 and payloads ascending from the lunar surface. The design, based on concepts from Tethers Unlimited including their "Lunavator" proposal, uses ballast and a movable mass to dynamically adjust rotation speed. This is analogous to an ice skater pulling in their arms to spin faster or extending them to slow down. The movable mass system also enables precise synchronization of the rotational period so that the tether tip points toward the Moon at perigee—the lowest point in the orbit.
The lunar environment offers significant advantages for initial tether deployment. Unlike Earth orbit, which is crowded with over 10,000 satellites, the Moon has only a handful of orbiting spacecraft, dramatically reducing collision risks. Additionally, the absence of atmospheric drag eliminates the need for continuous orbit maintenance and reboost operations.
The tether completes one orbit approximately every two hours, allowing for one exchange operation (one pickup and one release) at the low point of each revolution.
The most demanding engineering challenge, and therefore the primary technical risk, is "The Catch"—the moment when a payload trajectory must rendezvous with the tip of the rotating tether and successfully connect. This requires a high-precision positioning system around the Moon, analogous to GPS but potentially more accurate. Current or planned lunar positioning systems may not provide adequate precision.
Experimental catch mechanisms have been demonstrated to work within a few meters of separation, but these have not yet been tested at full scale with operational tethers. This represents a genuine technical risk that can only be retired through actual flight demonstrations.
A crucial advantage of tether systems is that they scale gracefully. The ratio of tether mass to payload mass remains roughly constant across different scales, which means starting with a smaller system doesn't incur fundamental efficiency penalties. An initial demonstration might use a 1,000-kilogram payload capacity with a 10,000-kilogram tether system—a relatively modest investment in space terms.
If successful, such a system could generate substantial returns immediately, easily funding rapid scale-up to larger capacities. If problems emerge, the modest initial investment limits financial exposure and makes iterative design improvements affordable. With Starship potentially achieving $25 per kilogram to LEO and using efficient ion drives for the slower transfer to lunar orbit, even a 10,000-kilogram tether demonstration becomes quite affordable.
The Earth-Moon L5 Lagrange point offers unique advantages for space-based computing infrastructure. Most discussions of orbital data centers focus on low sun-synchronous orbits, but L5 provides compelling benefits that may justify the higher energy cost to reach it.
Advantages of L5 over LEO for data centers:
Abundant cheap radiation shielding: Lunar regolith can be transported to L5 to provide radiation protection comparable to Earth surface levels, protecting both electronics and any human occupants.
Inexpensive solar panels: Panels manufactured on the Moon from local materials can be delivered to L5 far more cheaply than launching them from Earth.
Cheap radiators: Heat rejection is a major challenge for space data centers. Radiator panels produced from lunar materials provide cost-effective thermal management.
No atmospheric drag: Unlike LEO, L5 requires no station-keeping propellant to maintain orbit.
Unlimited thermal rejection: Massive radiator arrays can be deployed without concerns about orbital decay from increased drag.
No reentry pollution: There are no atmospheric pollution concerns from deorbited components.
No collision risks: L5 is far from the congested satellite bands around Earth, eliminating collision risks with operational or defunct satellites.
The Moon itself presents an equally compelling location for computing infrastructure, with some unique advantages that L5 cannot match.
The key insight is that solar power on the Moon can provide baseload power—something impossible with individual solar installations elsewhere. A power grid with solar panel arrays distributed around the lunar equator will always have approximately half of its panels illuminated, since half of the Moon is in daylight at any given moment. This provides continuous 24/7 power without requiring massive battery storage systems. On the Moon, solar power becomes steady, reliable baseload electricity.
If solar panels and transmission lines can be manufactured using ISRU from lunar materials, the levelized cost of electricity could be extraordinarily low. The proposal suggests that getting the equipment needed to manufacture solar panels and wiring to the Moon affordably is the key enabling step.
Both L5 and lunar surface locations can access abundant regolith for radiation shielding, allowing radiation levels inside data centers to match Earth-normal conditions. This protects sensitive electronics and makes any future human presence much more practical. Furthermore, ISRU-manufactured solar panels can be produced far more cheaply than panels launched from Earth, even accounting for Starship's dramatically reduced launch costs.
The tether orbits in a fixed plane relative to the stars, while the Moon rotates beneath it once every 28 days. This means the tether won't consistently pass over the same location on the lunar surface. To optimize the architecture, the tether's orbital plane should align with the Moon's orbital plane, which also aligns with L5's orbital plane. This configuration places the tether's orbit over the lunar equator.
Pickup and drop-off locations will therefore be distributed along the equator. This necessitates a transportation corridor—essentially a road—running along the equatorial region, with mobile equipment capable of positioning cargo for pickup and receiving descending payloads.
The transit time between low lunar orbit and L5 can extend up to 120 hours. With payloads launching every two hours, this results in approximately 60 payloads in transit in each direction at any given time, for a total of roughly 120 payloads simultaneously in flight. This establishes a minimum requirement of at least 120 cargo guidance containers.
Additional containers beyond this minimum would provide buffer capacity to allow time for loading, refueling, and maintenance. However, since these containers require only modest delta-V capability—they're primarily providing precision guidance rather than major propulsive maneuvers—they shouldn't be prohibitively expensive. As reusable capital equipment rather than expendable hardware, their cost is amortized over many missions.
Initially, operations would begin with a single container to limit risk exposure while validating the concept. The fleet would expand gradually as operational confidence grows. The containers may be light enough that returning empty units to Earth aboard Starship for loading and refurbishment is economically practical.
The Tethers Unlimited design concept has the tether's tip velocity exactly counteracting the tether's orbital velocity at release, resulting in a payload that is stationary relative to the lunar surface at approximately 1 kilometer altitude. From this point, the payload guidance module must provide the delta-V to descend safely, which amounts to approximately 58 meters per second.
Over time, as operational experience accumulates and confidence grows, this 1-kilometer handoff altitude might be reduced even further, decreasing the required delta-V budget for the payload guidance systems and improving overall efficiency.
Momentum exchange tethers require balanced traffic flows. For every kilogram going down, a kilogram must go up. Fortunately, the proposed architecture naturally creates such balanced flows.
On the lunar surface, operations would include mining, manufacturing of solar cells and metal products, and data center operations. At L5, the primary activity would be data center operations. Starship would carry lunar-bound cargo containers from Earth to L5, which would then be fed to the tether approximately every two hours.
Downbound cargo would consist of mining equipment, manufacturing machinery, and data center hardware. The Moon would send up an equal mass of solar panels, electrical wiring, thermal radiators, and radiation shielding material for use at the L5 facility. This creates a naturally balanced flow: Earth provides high-technology equipment that's difficult to manufacture in space, while the Moon provides mass-intensive structural materials that are expensive to launch from Earth but cheap to produce from lunar regolith.
Bitcoin mining operations offer an elegant solution for the early phases when traffic balancing might be challenging. Mining can start at small scale and expand gradually. More importantly, Bitcoin mining equipment can be readily moved between L5 and the Moon without disrupting operations—the modules are self-contained and location-independent. Some AI training workloads, by contrast, require substantial infrastructure before becoming viable and may not be interested until truly large-scale data centers are operational.
In the worst case, if other cargo flows prove insufficient, lunar regolith can always be sent up to L5 (useful for radiation shielding and other applications), while Bitcoin mining equipment or other modular computing hardware can be sent down to the Moon. This provides an absolute floor ensuring traffic can always be balanced, even if more economically optimal cargo flows temporarily falter.
The momentum exchange tether functions as a cargo multiplier. For every kilogram of mass that a rocket delivers to L5, the integrated system ultimately delivers nearly two kilograms of useful material to their final destinations. This is possible because the tether recycles the kinetic energy from each kilogram descending from L5 to the Moon and uses that energy to lift a matching kilogram from the Moon to L5. No additional propellant is consumed for this exchange—it's pure momentum transfer enabled by the tether system.
The first several payload deliveries would be descent-only operations, not true momentum exchanges, since nothing would yet be on the lunar surface ready for uplift. To recover the momentum imparted during these initial one-way deliveries, the tether system would use onboard ion thrusters. This process takes considerable time, so the initial cadence might be one payload per month rather than one every two hours, depending on the solar power available and the thrust capacity installed on the tether.
Ideally, the very first cargo delivery would include a simple machine capable of filling bags or containers with the precisely correct amount of regolith. After this initial delivery, traffic could immediately become bidirectional, with regolith bags providing the uplift mass to balance subsequent downbound cargo.
Initially, Starship would serve as the primary means of transporting cargo to L5. This likely requires refueling in LEO using propellant from approximately three additional Starship tanker flights. The total cost therefore represents four Starship launches to LEO, making the effective cost to L5 roughly four times the cost to LEO. If Starship achieves its target cost of $25 per kilogram to LEO, then delivery to L5 would cost approximately $100 per kilogram.
Longer term, a dedicated high-ISP space tug operating between LEO and L5 would be far more efficient. This could employ nuclear thermal propulsion, nuclear electric propulsion, or advanced concepts like fusion-enhanced thrusters. The key requirements are high specific impulse (ISP) and sufficient power levels to complete the round trip from LEO to L5 and back to LEO in a reasonable timeframe—perhaps two weeks to one month.
Such a tug would consume less than half its mass as propellant for the round trip. Instead of Starship's effective cost of four times LEO prices to reach L5, a nuclear tug could reduce this to approximately twice LEO costs. Combined with the tether's cargo multiplier effect (which doubles useful payload from L5 to the Moon with no additional propellant), a nuclear tug would enable lunar delivery costs approximately equal to LEO costs. This would be remarkable.
An even more tantalizing possibility emerges with sufficiently high-ISP propulsion. The space tug's reaction mass could potentially be supplied from the Moon rather than from Earth. If the tug achieves high enough ISP that it consumes less propellant mass than the cargo mass delivered to the Moon—even when refueling at L5 instead of LEO—this architecture becomes viable. It requires higher ISP because the tug must now travel from L5 to LEO with full propellant tanks, which increases fuel consumption. However, with advanced propulsion systems like nuclear electric drives with fusion-enhanced thrusters, this becomes feasible. In this configuration, Starship would only transport cargo, not propellant, dramatically simplifying Earth launch operations.
To properly evaluate the tether system's economic advantages, we can compare it to direct Starship lunar landings. This analysis involves significant uncertainties, but provides useful order-of-magnitude insight.
A Starship lunar landing mission likely requires approximately ten refueling flights in LEO to fill the ship's propellant tanks (estimates vary considerably). To land on the Moon and retain sufficient propellant reserves for takeoff and Earth return, Starship can only carry about half its nominal cargo capacity—approximately 100 tons rather than 200 tons. This effectively means eleven Starship launches deliver 100 tons to the lunar surface, or equivalently, twenty-two Starship launches per 200 tons of cargo.
At a target cost of $25 per kilogram to LEO, this translates to $550 per kilogram to the lunar surface via direct Starship landing (with substantial uncertainty in these figures). The tether-based system, by comparison, could achieve approximately $50 per kilogram—roughly eleven times more cost-effective.
This analysis suggests that lunar ISRU, a lunar momentum exchange tether, and a data center at L5 create a powerful synergy that could fundamentally transform lunar economics.
If this project moves forward, the scale of investment involved justifies significant research into advanced tether materials. Current tether ropes would work, but substantial funding could likely yield significantly improved materials with better strength-to-weight ratios, greater longevity, and improved resistance to the space environment.
Better ropes would enable larger safety margins, and safety is paramount in any system involving momentum exchange with occupied facilities. The potential economic returns from this system could easily justify a serious materials development program.
Starship is poised to make space access far more affordable than ever before. Many concepts that couldn't attract funding or achieve viability in the past will suddenly become practical. If the electricity cost projections work out favorably, space-based data centers become not just feasible but economically compelling. We're likely to see hundreds of billions of dollars of economic activity in space within the next few decades.
This will justify a level of private space development unprecedented in history. Technologies like space nuclear power, fusion-enhanced thrusters, space tethers, and ISRU—which previously received only modest government research funding—will soon attract large-scale private investment because they'll solve real, profitable problems. The rate of advancement in these fields will accelerate dramatically.
This proposal presents an integrated architecture that could reduce lunar delivery costs to approximately the same level as LEO launch costs. It doesn't require exotic materials, breakthrough physics, or technological magic. It does require solid engineering work and operational validation.
The fundamental components are:
• High-ISP space tugs (likely nuclear-powered) for efficient LEO-to-L5 transport
• A momentum exchange tether system around the Moon
• In-Situ Resource Utilization on the lunar surface
• Computing facilities at L5 and on the Moon to drive economic demand
The potential profits from space-based computing operations can justify the substantial development work needed to realize this vision. As launch costs continue to decline and space-based electricity becomes genuinely cheap, the economic logic becomes increasingly compelling. We may be on the cusp of a genuine transformation in how humanity accesses and utilizes the Moon.
Space tug using fission and fusion - Vince Cate
Fission-Powered Fusion Rocket - Alex Taits
Helicity Space - Fusion enhanced thruster technology
RocketStar - FireStar™ Fusion Drive
Pulsed Fission Fusion (PUFF) Propulsion System - NASA Technical Report
Rob Adams: Pulsed Fission Fusion - Blue Marble Week presentation
Pulsed plasma rocket - ScienceDirect
Why Nuclear Rockets Are Going To Change Spaceflight - Scott Manley
Project Athena - Nuclear Electric Propulsion priorities - Isaacman
Space Tether - Grokipedia
Momentum Exchange Tether - Grokipedia
NIAC Tether Studies - Final Report
Why Everyone Is Talking About Data Centers In Space - Scott Manley
Economics of Orbital vs Terrestrial Data Centers - Online calculator by Andrew McCalip
Space-Based Data Center - Grokipedia
Nuclear Pulse Propulsion - Wikipedia
Nuclear Thermal Rocket - Grokipedia
Nuclear Electric Rocket - Grokipedia
Magnetoplasmadynamic Thruster - Grokipedia
Direct Fusion Drive - Grokipedia
Fusion Rocket - Grokipedia
Pulsed Plasma Thruster - Grokipedia
Lagrange Point - Grokipedia
Space Tug - Grokipedia
In Situ Resource Utilization (ISRU) - Grokipedia
Optimus Robot - Grokipedia
Rocket Equation Calculator - Aerovia