How to allocate orbits around the sun

Imagine infrastructure around the Sun that isn’t the occasional satellite but something much larger: enough solar captors (essentially solar panels) to catch a meaningful fraction of the Sun’s light. This is the rough shape of a Dyson swarm. To many it sounds like science fiction, but the underlying engineering isn’t exotic. Solar panels, orbital mechanics, and launch capacity, scaled up a lot. If humanity reaches that scale, the question of who builds where around the Sun becomes one of the most consequential resource allocations in history.

The Sun matters because it is a fundamental bottleneck. Large-scale expansion through space requires energy and matter, and the Sun holds the overwhelming majority of both in our solar system: roughly 99.8% of its mass, and more energy output in a second than Earth consumes in a year. Other bottlenecks could appear — conflicts between actors, strong regulations — but given favourable conditions for expansion, energy and material constraints are never far behind. So whoever has preferential access to the Sun is positioned to expand at the greatest scale.

The reason to think about this now is that “later” may arrive sooner than most expect. As AI capabilities continue to improve, we seem likely to experience an intelligence explosion which could introduce a century’s worth of new or improved technologies within just a few years, potentially facilitating much quicker expansion to and across space. A subsequent industrial explosion would then allow increasingly fast scaling of the space industry as self-replicating or largely autonomous factories build more factories to build more solar captors. Under these conditions, it might not be very long until an actor could build a Dyson swarm and take control of the solar system along with the long-term trajectory of sentient life. Alternatively, depending on how AI develops, large-scale deals about the solar system could be struck quite early, before the questions in this document have been worked through carefully.

Some industry actors seem to be heading in the general direction. SpaceX and Blue Origin have both announced Earth-orbiting data centers, and Elon Musk recently told a SpaceX all-hands that the goal is to “extend the light of consciousness to the stars.” Solar captors orbiting the Sun would be a key step in this mission, and none of the technology requires physics we don’t already understand. With the technological uplift from intelligence and industrial explosions — which SpaceX may be in a particularly good position to capitalise on — the first solar captors could plausibly be deployed within the next decade, setting early precedents that shape the regime that follows.

Beyond timing, there’s a structural reason allocation matters: it helps avoid issues that may arise from market forces alone. Orbits near the ecliptic (the flat disc defined by Earth’s own orbit, which most of the solar system rides on) and close to 1 AU (the Earth–Sun distance, roughly 150 million km) are the most commercially attractive¹: easier to reach and operate from. Absent governance efforts, a first mover is incentivized to pack captors into this commercially attractive band, which gives leverage over how much sunlight reaches Earth, at a scale far out of proportion to the fraction of total solar output they actually capture.

The general pattern is broader: concentration in particular regions of orbital space tends to confer strategic leverage that a more evenly distributed allocation would not. By default, the actor that can scale fastest and dominate others is likely to take these regions, accruing an increasing fraction of power over time. This means that how rules are set, if any are set at all, will shape which kinds of actors — whether states, companies, alliances, or something else — are favoured to emerge and succeed. Early rules in novel domains tend to stick, and how we start expanding across the stars, if we make it that far, could play an outsized role in how the future goes for all sentient beings who come to exist in it.

So it seems worth looking now for allocation schemes that more plausibly lead to good futures rather than bad ones. This is a complicated problem, with lots of uncertainties on what "good" and "bad" might look like. To start somewhere, I will posit that important properties² which seem worth aiming at include: preventing undue concentration of power, transparency about who holds what, and the ability to revise the rules as we learn more. Finally, any scheme has to be agreeable enough to the actors involved, since a framework no one adopts has no effect, and one participants walk away from collapses into a free-for-all.

The rest of this document attempts to start a deeper exploration into how to allocate orbits. §02 and §03 establish what orbits are and what allocation schemes look like; §04 examines the factors that make certain orbits more valuable; and §05 connects those factors to the governance properties above.

1. [To be added: motivation for this choice of properties.] ↑
2. See Lagrange Point 1, a gravitationally convenient position between Earth and the Sun where infrastructure is relatively easy to park and operate. ↑

Any orbit around the Sun can be uniquely defined with six numbers. There are many equivalent ways to choose them, but for intuitive interpretation we use the classical orbital elements.^↗ The first two describe the shape of the orbit. The next three describe how the orbital plane is oriented in space. The last describes where on the orbit the captor happens to be at a given instant. Use the visualization below to build an intuition for what each parameter does.

Each of these six numbers can be “allocated” to an actor that wants to place captors, or other infrastructure, in orbit around the Sun. How that allocation is structured carries significant implications, which the next sections explore.

With a vocabulary for describing individual orbits, we can ask how a governance framework might assign them. The visualizations below build intuition for each orbital dimension in turn, then combine them to show examples of more complex orbital allocation structures. There are many more combinations that have yet to be explored! Three illustrative actors — A, B, C — are used throughout.

Radius (a): fixed shell vs variable

If eccentricity is zero, every orbit follows a circular path. A set of circular orbits at the same radius share a shell and exactly two crossing points. This means captors on a single shell require time-coordinated collision avoidance, placing a practical ceiling on how densely a shell can be packed. Allowing variable radius removes the constraint: captors on different shells won't be able to collide.

For the remaining visualizations, variable radius is assumed throughout. As the geometry above shows, single-shell packing is collision-constrained in a way that multi-shell allocations are not — making variable radius a near prerequisite for any full-scale satellite swarms. All but the ν visualization assume that all 360° of ν are allocated to a single actor, which makes orbits appear as bands. For any of these visualizations, you could (and likely should!) allocate ν across multiple actors.

Inclination (i): The North–South allocation

Starting from the ecliptic (i = 0°), increasing the inclination range adds new bands, each allocated to the next actor in sequence.

Longitude of ascending node (Ω): the cocoon allocation

Ω rotates an orbital band around the Sun’s axis at a given inclination. Set an inclination, then increase the orbit count to see the structure that emerges.

Eccentricity (e) and argument of periapsis (ω): orbit shape and orientation

Eccentricity (e) stretches a circular orbit into an ellipse; as e increases the Sun shifts toward one focus and captors spend more time near the far end of their path. Argument of periapsis (ω) rotates the ellipse within its orbital plane. All orbits here share the same inclination; each new orbit adds 15° to ω. The periapsis marker shows where each orbit comes closest to the Sun — at e = 0 the markers all stack (periapsis is undefined for a circle) and only spread into a rosette as eccentricity increases.

True anomaly (ν): dividing a single orbit between actors

Even a single orbital path can host many captors, staggered in position along it. The 360° of each orbit are divided into 6° segments below; as the slider fills them in, three actors share each path in sequence.

Unstructured allocation: fully random

As a baseline, consider what happens when five orbital elements — radius (a), eccentricity (e), inclination (i), longitude of ascending node (Ω), and argument of periapsis (ω) — are drawn independently at random. No coordination is imposed; each orbit is assigned without regard to the others.

Structured allocation: Fibonacci lattice (a, i, Ω)

A Fibonacci (golden-angle) lattice spreads orbital planes near-uniformly across all directions, which is a property not guaranteed by random assignment. With variable radius, this distributes screening locations throughout the swarm rather than concentrating them. This follows from the structure itself, without requiring explicit coordination between actors. It may not be the easiest framework to establish or amend over time.

Structured allocation: Ω then i

There is a physical reason to prefer filling Ω values before stepping to a new inclination. Changing Ω at a fixed inclination costs almost no additional delta-v as you can achieve it simply by waiting for the right launch window as the Sun’s position shifts. Changing inclination requires an actual burn. So the lowest-cost allocation strategy is to exhaust feasible Ω slots at one inclination before paying the delta-v price to move to the next. Below, each band of four orbits shares an inclination; Ω steps through 0°, 90°, 180°, and 270° before inclination increases by 10°.

While it is hard to predict how technologies develop over time, we can try to identify the factors that matter when choosing an allocation framework. Choosing a sufficiently amendable framework means the allocation can be updated to reflect whichever factors are most salient given current conditions. You can find the full catalogue of currently identified factors with brief descriptions in the appendix. In this section, I will mention and motivate the few factors that I think will be particularly important.

Energy availability and screening

Solar power per unit area scales as 1/r², so an orbit at 0.1 AU captures roughly 100 times more energy than one at 1 AU — meaning frameworks counted in captor number can be profoundly unequal in energy value. Screening compounds this: inner captors periodically shadow outer ones, creating structural pressure toward the innermost feasible orbits. This pressure further exacerbates screening because fewer captors are needed to fill a shell at a smaller orbital radius; the space fills up more quickly.

Captor–Earth screening adds a further political dimension: dense swarm populations transiting directly between the Sun and Earth reduce the amount of sunlight reaching Earth, and at sufficient density this would become a measurable climate intervention well before full coverage is achieved.

Delta-v cost

Delta-v is effectively the cost to reach and maintain any given orbit. Unless you wait for very specific launch windows, delta-v costs tend to increase as captors are put in orbits closer to the Sun or at higher inclinations. While technology will likely improve over time and reduce delta-v for a given orbit, reaching orbits close to the Sun and at high inclinations can increase delta-v by roughly 1–2 orders of magnitude, partially offsetting early technological gains.

As a result, the cheapest orbits to reach at any distance would be those close to the ecliptic plane (0° inclination), which is why we may expect near-ecliptic orbits to fill up quickly. However, a high concentration of captors in these orbits produces a higher captor-captor screening fraction, reducing energy availability and creating an incentive for actors to move to another Ω, and subsequently pay a higher delta-v cost to reach higher inclination orbits.

A final consideration is that the order in which you execute orbital manoeuvres matters. Say a captor launched from Earth needs to reach an orbital radius of 0.05 AU and an inclination of 10°. Changing inclination at Earth’s orbital radius before spiralling inward is roughly 4.5 times cheaper than doing so after arriving at 0.05 AU, because higher orbital speeds at smaller radii make inclination changes more costly. This raises the expected cost of amending an allocated orbit — the closer to the Sun, the more expensive any inclination adjustment becomes.

Actor interactions

Different allocation schemes create different inter-actor dynamics. An allocation that assigns all 360° of true anomaly in a given orbit to one actor makes it easier to preferentially damage that actor’s assets. Mixing actors along the same orbital path raises the cost of targeted interference, since destructive actions risk friendly fire — debris from destroying one actor’s captor is liable to damage others’ nearby. This dynamic could serve as a deterrent to aggressive actions.

These interspersed orbits could also make it easier to monitor neighbouring actors, improving overall transparency depending on the governance system in place.

If we are looking for allocations that have the four properties of resisting concentration of power, amendability, transparency, and agreeableness (to participating actors and external stakeholders alike), how should we think about the orbital parameters to allocate in light of the factors that seem relevant?

True anomaly (ν)

Allocating true anomaly among various actors for any given orbital band seems important to resist concentration of power. If this is not done during the early captor buildout, a first mover would gain preferential access to cheap orbits (e.g. 0° inclination), forcing subsequent actors to pay higher delta-v costs (e.g. 1° inclination). Allocating ν could also help increase deterrence, and lead to improved transparency of the overall system through increases in monitoring between actors.

Semi-major axis / radius (a)

Actors are incentivised to research and employ technologies that allow them to orbit closer to the Sun because of the increased energy availability. To prevent collisions and excessive screening — unless or until statites¹ can be employed — radius should be allocated for the sake of agreeableness. Orbits at smaller radii cost more delta-v to amend, all else equal, which could be important to consider until improved in-space propulsion technology is available.

Inclination (i)

Inclination requires more delta-v to achieve than Ω, but will need to be balanced with Ω because successive inclination bands at a fixed Ω increasingly screen each other. Inclination allocation is necessary to complete a full Dyson sphere. Together with ν allocation, inclination requirements address the concentration risk described in §01: without them, early actors could pack the near-ecliptic band and gain disproportionate leverage over how much sunlight reaches Earth.

Longitude of ascending node (Ω)

Allocating Ω is not strictly necessary to complete a full Dyson sphere. However, for any given orbit, shifting Ω is an almost free way to reduce screening, compared to the delta-v cost of an inclination change. In fact, no deliberate waiting is required: sequential construction naturally produces varied Ω values over time as Earth moves around the Sun. This means that actors, particularly near the beginning of Dyson swarm construction, would prefer Ω allocations rather than increasing inclination. Ω and inclination allocations therefore need to be balanced. The Ω-then-i allocation in §03 is one example of how this may look.

Eccentricity (e)

Given a specific orbital radius, energy availability is maximised by a circular orbit at that radius, implying that actors would want an eccentricity of 0. While other factors like line-of-sight, delta-latency, and thermal cycling could further support e = 0 being desirable, there are some conditions where myopic optimisation for energy availability could lead actors to desire eccentric orbits. For example, an actor allocated a fixed average orbital radius might request a nonzero eccentricity so their periapsis dips closer to the Sun, achieving greater energy availability there while technically averaging to the allocated radius. These dynamics need further exploration.

Argument of periapsis (ω)

Given an eccentricity of 0, argument of periapsis is undefined and need not be allocated.

Allocating and tracking solar infrastructure using the above parameters is necessary, but not sufficient, for transparency. A governance regime also needs registries, monitoring infrastructure, and enforcement mechanisms that go beyond which orbits are assigned to whom.

Finally, it’s important to note that the parameters that matter most shift as the swarm grows, and as technology changes which orbits are feasible or desirable. Under the condition of an industrial explosion in space, the completion of a full Dyson swarm could follow an exponential trajectory, with a large majority of captor construction occurring within the final few percent of its total construction time. Given this, it seems wise to focus more on the allocations and factors that seem important in early stages of solar energy capture, and punt those that are not into the future, provided the allocation chosen is amendable. This asymmetry also means that the governance regime established before the exponential phase begins matters disproportionately as course-correction becomes increasingly difficult during the steep phase of the exponential swarm build-out.

1. Statites are spacecraft that use radiation pressure from sunlight to hold a fixed position relative to the Sun without an orbit, potentially eliminating collision risk and enabling arbitrary positioning. ↑

Thanks to my mentor Jordan Stone for conversations and feedback that shaped this piece.

Thoughts on the content, errors spotted, or questions about the framework — all welcome.

Thanks — feedback received.

Energy availability

• Solar flux — Power received per unit area scales as 1/r² with distance from the Sun, making inner orbits dramatically more productive per captor.

Screening

• Captor–captor screening — A captor on an inner orbit periodically blocks sunlight from outer captors at the two points where their paths cross.
• Back-side screening — Captors at larger radii can intercept energy being radiated or transmitted outward from captors on inner orbits facing away from the Sun.
• Captor–Earth screening — Dense captor populations transiting the Sun–Earth line progressively reduce the sunlight reaching Earth; at sufficient density this becomes a climate-relevant intervention.

Captor lifespan risks

• Thermal cycling — Repeated or large temperature swings from eccentric orbits or sustained screening accelerate material fatigue and degradation.
• Radiation and solar weather exposure — CMEs, particle events, and solar winds erode hardware over time; risk varies by orbital location and likely tracks the ~11-year solar cycle.
• Natural obsolescence — Components have finite operational lifetimes independent of the space environment.
• Space debris and dust — Collision and abrasion risk is higher near the ecliptic plane and in orbital zones with higher particle density, such as near the asteroid belt.

End-to-end captor costs

• Delta-v cost — The total propellant budget to reach, enter, and maintain an orbit; dominated by launch site conditions, radial distance, inclination change, and ongoing station-keeping against perturbations.
• Repair and maintenance — Servicing captors would require dedicated infrastructure and delta-v; it is currently unclear when repair or replacement is more cost-effective.

Quality of allocated orbit

• Line-of-sight with Earth — Direct visibility enables communications and potential energy transfer; frequency and duration depend on semi-major axis and inclination, and vary over the orbital period.
• Communication latency — Light-travel time to 1 AU is about 8.3 minutes; the variation in latency over an orbit (delta-latency) may matter more than the absolute figure for most applications.
• EM noise environment — Dense solar emission across the spectrum near the Sun may limit available transmission bands, particularly for captors at small radii or in angular proximity to the Sun as seen from Earth.

Actor interactions

• Transmission interference — Dense near-Sun populations all transmitting toward Earth may require frequency coordination analogous to geostationary satellite management.
• Collision risk — Increases with captor density, particularly during orbit entry and amendment, though the vastness of even inner solar space could keep absolute probability low.
• Waste heat — If captor density is particularly high in a given area, dissipating waste heat within a dense swarm could become a physical constraint on total throughput.
• Monitoring — Some allocation schemes keep different actors’ captors interspersed, enabling more inter-actor observation. Other allocations concentrate actors in bands, providing fewer opportunities to monitor each other.
• Targeting vulnerability — Band-based allocations make it easier to preferentially damage one actor’s assets; mixed-actor orbits raise the cost of targeted interference by increasing the chance that destructive actions hurt everyone.
• Kessler-like syndrome — Debris accumulation in heavily used orbital zones could render them unusable, especially at small radii where the orbital shell is physically narrowest and replacement rates highest.