The rules for a cleaner orbit, explained

This Rulebook defines how a mission's debris exposure may be described (SDER) and how a verified reduction in long-lived debris risk may be recorded as a claim (SDC). Its purpose is to make the cost of leaving long-lived risk in orbit visible, and to give States, insurers and other buyers a common unit for paying for its reduction. It covers objects in Earth orbit and the operators, servicers and launching States associated with them. The activities in scope are launches, operations, end-of-life disposal, passivation and active removal.

Nothing in this Rulebook overrides the jurisdiction and control of the State of registry, the liability regime, or national licensing law. The administrator publishes rules, verifies outcomes and keeps the record. It holds no enforcement power.

Participation is voluntary unless a licensing authority makes it a condition. Only States and their authorities may impose sanctions, such as fines, licence conditions, suspension or refusal of a licence. The administrator may refuse to record a claim, revoke a recorded unit (C10), suspend a verifier's accreditation (D1), and publish these decisions. It reports suspected non-compliance to the relevant licensing authority.

An SDER assessment describes a mission's exposure across its lifecycle under a published method version. It is not a ranking of operators. No league tables are published until coverage, peer review, correction and appeal processes are working.

A method version lists its inputs. The minimum set covers:

• orbital regime and persistence;
• mass and cross-section;
• disposal plan and reliability;
• passivation;
• maneuverability and coordination;
• conjunction exposure.

Every input carries an evidence class and a source reference. A component takes the class of its weakest input, and a weaker class widens the component's uncertainty interval.

A missing input is never replaced by a favourable value. It is scored at its worst case and named in the output. The result is indicative when nothing is missing and provisional when the missing weight is at or below a published threshold. Above that threshold the result is withheld.

Every result gives a point estimate and an interval. A band (low, medium or high) is reported together with whether the interval lies entirely inside it.

Weights, tables and thresholds are published as a versioned method file. Each assessment records the method's identifier and its SHA-256. A new version never changes the meaning of a past assessment.

An assessed operator receives the inputs, the evidence classes and the rationale for each component. The operator may submit corrected evidence, and may appeal to a panel that took no part in the assessment.

If a scheme reports an operator-level figure, it is computed from the operator's mission assessments under one method version, with the number of missions and the share withheld shown next to it. It is published only once B1's conditions are met.

An SDER assessment may be used for disclosure, to adjust an obligation (F4), to inform an orbital-use fee (F1), or as an insurance or procurement criterion (F1). Each use names the method version and the band certainty (B5). A withheld result cannot be used to reduce an obligation.

One SDC unit represents a verified reduction of ΔR₀ in the risk proxy over a stated horizon, relative to a stated baseline. The number of units for an action is calculated from the conservative bound of ΔR, meaning the lowest reduction consistent with the evidence, rounded down.

Only the following classes are eligible during a pilot:

Baselines are versioned by jurisdiction and by date. Doing what the law requires earns no units. Only performance beyond the applicable requirement is credited.

Disposal credits apply only to performance beyond the baseline. Legacy-removal credits are restricted to objects launched before the scheme, or to removals by an entity other than the original operator acting under explicit consent. Newly launched objects become eligible for disposal credits only after a waiting period.

A claim lists references to its evidence. Evidence covered by confidentiality may be held by the verifier, but the registry publishes a commitment to it (a hash) and the verifier's finding.

An action on an object outside the claimant's domestic jurisdiction is eligible only with documented consent from the State of registry. Before recognition, the claimant shows how liability is allocated between servicer, State of registry and licensing State.

A claim moves only along these transitions:

• claimed → under_review or rejected
• under_review → verified or rejected
• verified → issued or revoked
• issued → retired or revoked
• retired → revoked
• rejected and revoked are final.

Only one non-void claim may exist for a given (object identifier, outcome type). Within one action, an object may appear only once. Credits go to a single accountable entity, typically the operator under the authorising State. A service provider may receive the economic value by contract, but the registry still records only one issuance for the physical event.

Retirement is one-way and records who retired the unit and for what purpose (an obligation, named; or a voluntary claim).

reject and revoke require a public reason. If a recorded outcome proves false, for example because an object reported as reentered is re-acquired by tracking, the claim is revoked and the revocation is published. A claimant found to have made a knowingly false claim is barred from submitting claims for a published period, and the administrator informs its licensing authority. A verifier that confirmed it without due care loses accreditation (D1).

An action is eligible only if the objects it concerns can be reliably tracked before and after, by at least one data provider independent of the claimant. The administrator publishes the size threshold for each orbital regime and widens the scope as tracking improves (D6).

No unit is issued for an action that itself creates undue risk, such as deliberate fragmentation, or an action carried out in breach of flight-safety rules or without the coordination a licensing authority requires. If an action produces debris, that debris is counted against ΔR.

A claimant files a plan for any action it intends to claim, before the action, with its baseline, method version and evidence plan. Actions not notified in advance may be claimed only if the evidence required by C2 is still complete, and they are marked as such.

Units are counted in ΔR, not per object or per kilogram removed.

Verifiers are accredited against published criteria: technical competence, independence and data handling. A verifier must not be the claimant and must declare conflicts. The registry refuses a conflicted assignment. Only the assigned verifier may record verification. The administrator publishes the list of accredited verifiers and data providers, and may add, suspend or remove them.

Wherever possible, an outcome is confirmed by at least two independent data sources, for example two tracking networks, or tracking plus operator telemetry. If the sources disagree, or the confirmation threshold is not met, a third-party audit starts automatically and no unit is issued until it ends.

The administrator rotates verifiers, peer-reviews a sample of verifications and re-audits a random sample. Part of the verification fees may fund this work.

Disputes go through three tiers:

1. Informal resolution, facilitated by the administrator: sharing data, correcting inputs, recalculating.
2. A dispute panel combining technical and legal expertise, under published procedures and timelines. Every party affected has the right to be heard, and decisions rest on the evidence and give reasons.
3. External recourse. Nothing in this Rulebook prevents a party from going to national courts or diplomatic channels. A court decision that bears on a recorded outcome is applied in the registry.

When disputes show that a rule is systematically ambiguous, the rule is clarified in the next method cycle.

An independent auditor, not the administrator, audits the scheme each year and publishes the results:

• data audit: a sample of issuances checked against raw tracking data and mission logs, and a sample of SDER inputs checked against their sources;
• financial audit: all fees, penalties received and ESAF flows (H7);
• procedural audit: that rule and method changes were made as decided (G1), and that every manual intervention was logged with a reason.

The administrator reviews the evidence requirements (C2), trackability thresholds (C11) and method versions (B6) at least once a year, through the change process in G1.

Where an obligation exists (Part F), the administrator compares each participant's retirements with its obligation once per period and publishes the result. A shortfall is handled in steps: notice, then a published grace period, then a report to the licensing authority, which decides any sanction (A6).

The registry records every event: submission, verifier assignment, verification, rejection, issuance, retirement and revocation. Each record carries a sequence number, a timestamp, the actor and the payload.

The registry is an append-only log in which each entry carries the hash of the previous one. When the log is opened, it is replayed and re-checked against these rules. The head hash is published on a schedule in at least one location the administrator does not control.

The administrator publishes aggregate outcomes, including total ΔR retired. It also publishes methods, anonymised verification statistics (time, cost, dispute rate) and uncertainty ranges for key parameters. Sensitive operational details are not published.

Each year the administrator publishes an orbital sustainability report covering: ΔR recorded and retired by action class and regime; claims rejected and revoked, with reasons; audit findings (D5); compliance review results (D7); ESAF accounts and grants (H7); and progress on the indicators in Part J.

This Rulebook does not presuppose a market. Recorded outcomes may serve any of the following:

Whatever instrument is chosen, a published share of the scheme's revenue goes to the Equitable Space Access Fund (Part H).

A jurisdiction that sets an obligation publishes: who is covered; the unit per object launched or per unit of SDER exposure; the period; how initial allocations, if any, are made; and an aggregate ceiling. The ceiling is set so that, across covered operators, long-lived risk does not grow. A stricter ambition, in which each new object is matched by a recorded reduction ("one-up, one-down"), is one possible end point.

An obligation may be adjusted by SDER band. Missions in the low band with an interval wholly inside it may have their obligation reduced. Missions in the high band may have it increased, for example to two units where one would otherwise apply. Missions in the middle band, or whose interval crosses a band boundary, keep the nominal obligation. The adjustment schedule is published with the method version.

Where a licensing authority imposes a financial penalty for a shortfall, it is invited to direct a published share of it to the ESAF (H2). The administrator itself imposes no financial penalties (A6).

Changes follow GOVERNANCE.md:

• a proposal with rationale and sources;
• at least 30 days of comment for substantive changes;
• a published decision with reasons;
• versioned effect, so a change never applies retroactively.

An independent expert group is drawn from operators, insurers, regulators and independent researchers, with conflicts declared. It proposes method updates. Voting rules prevent any one class of participant from dominating.

The originator and any affiliated organisation may implement the method. They may not assess or certify their own missions or claims, chair the review group permanently, hold a seat on the ESAF board, or veto a method.

The administrator:

• keeps this Rulebook and its annexes (G1);
• accredits verifiers and data providers (D1);
• keeps the registry (Part E);
• runs compliance reviews where obligations exist (D7);
• runs tier 1 of dispute resolution and convenes tier 2 (D4);
• publishes the annual report (E4).

It does not set obligations or quotas; licensing authorities do (F3).

The administrator reports annually to the States that recognise the scheme, and offers its reports to COPUOS through a participating State or observer organisation.

Governance widens as participation grows. When published milestones are met, seats in the review group and decision rights pass from founding participants to broader constituencies. The milestones concern the number of participating States and operators, the share of participants that are emerging space actors, and years of operation.

The following can be changed only by a supermajority set out in GOVERNANCE.md, after at least 90 days of comment: A4 (relationship to law), C3 (baselines), C6 (consent), G3 (originator), and the existence of the ESAF and its beneficiary floor (H2, H6).

The ESAF exists so that debris rules do not become a barrier to entry for States and organisations with fewer resources, and so that they can take part in tracking, verification and remediation. It follows three principles:

• Outer space is "the province of all mankind" and is free for exploration and use by all States (OST Art. I).
• Responsibility is proportional: those who place more long-lived risk in orbit contribute more.
• Compliance should build capacity, not only cost money.

The ESAF receives a published share of each revenue stream the scheme or a recognising jurisdiction creates:

• proceeds of any initial allocation or auction of obligations (F3);
• verification and registry fees;
• penalties directed to it (F5) and forfeited bonds (F1);
• fees on transfers, if transfer is ever permitted (F1).

It may also receive voluntary contributions from States, operators and foundations, and co-financing from development banks. No contribution buys influence over grant decisions.

An ESAF board governs the fund, separately from issuance decisions. It has equal numbers of seats for:

• emerging space actors (A2), with regional balance;
• contributors.

It also has non-voting seats for independent experts, for example from UN development or space bodies. The chair rotates between the two groups each year. Decisions are by consensus where possible; otherwise by a two-thirds majority that includes a majority of each group. The originator holds no seat (G3).

The board publishes calls for proposals and its selection criteria. The criteria cover: the expected effect on orbital safety or on the applicant's ability to take part; the applicant's capacity to deliver; value for money; and regional balance. Every decision is published with reasons. Applicants may ask for a review once.

At least 50% of disbursements each year go to projects led by, or directly benefiting, emerging space actors.

The ESAF publishes annual accounts, every grant with its amount and purpose, and an impact report against stated objectives. An independent audit of its finances is carried out each year (D5) and published.

The board holds an annual open forum, where possible alongside a COPUOS session or the International Astronautical Congress. Funded projects report there and contributors announce pledges.

When a proposed rule change (G1) would raise obligations, the board is consulted on its effect on emerging space actors and may propose transitional support. The board itself cannot change H2's shares or H6's floor; those follow G1 and G7.

The scheme is introduced in four phases. A phase begins only when the evidence gates of the previous phase are met and published. The phases are not tied to dates.

Jurisdictions that implement this Rulebook recognise units issued under each other's implementation, provided both use the same unit definition (C1), publish their baselines (C3) and share one registry or linked registries that enforce uniqueness (C8). Each jurisdiction publishes a conformance statement.

If removal capacity lags behind obligations, authorities may phase obligations in more slowly, and the ESAF may co-fund removal missions. If recorded units greatly exceed obligations, authorities may tighten obligations sooner. If transfer is permitted, any price-stabilisation measure (for example a reserve) is published in advance with its trigger.

If the Part J indicators show that long-lived risk is falling and disposal is routine, the review group may propose moving the scheme to a maintenance mode with lighter reporting. This change follows G1.

The administrator tracks and publishes, each year:

• total ΔR recorded and retired, by action class and regime;
• total R in each target shell, from an open catalogue, and its trend;
• verification cost, time and dispute rate (E3);
• the share of missions with indicative, provisional and withheld ratings;
• the number of participating States and operators, and the share that are emerging space actors;
• ESAF income, disbursements, and the beneficiary share (H6).

Every five years, an independent review asks whether the scheme has reduced long-lived risk relative to what would likely have happened without it, and whether its costs fall fairly. The review is published with the data it used.