100,000 tonnes are launched to orbit in a year

Where things stand

My own count from GCAT data. I downloaded the GCAT object catalogs, data update 2026 Oct 11, from https://planet4589.org/space/gcat/tsv/cat/. I used satcat.tsv, satcat070k, satcat100k, satcat270k, auxcat, csocat and tmpcat. I left out satcat700M, because it holds alias copies of the same objects, and deepcat, because it repeats objects already listed. I summed the Mass field for every non-debris object whose launch tag is a normal orbital designation (YYYY‑NNN).

Results:

Year	Mass inserted to orbit
2023	~2,300 t
2024	~3,109 t
2025	~4,217 t (the record)
2026, through Oct 10	~3,166 t

• At the same date in 2025 the running total was 3,072 t, so 2026 will finish around 4,300–4,600 t.
• Before 2025, the highest years were the Shuttle-era peaks of about 2,000–2,400 t.
• The 100,000 t threshold is therefore about 23 times today's level.
• The resolving sources agree on the scale. BryceTech's 2025 payload-only figure was about 2,700 t, and SpaceX's prospectus reports 2,213 t of SpaceX payload in 2025.

How GCAT counts Starship.

• Flight 14 (Sept 28, 2026) was Starship's first orbital flight. GCAT records the ship as S100881 with a mass of 120,000 kg, plus 26 Starlink V3 satellites at 2,000 kg each, for 172 t in total.
• The Super Heavy booster does not reach orbit, so it is not counted.
• An operational Starship flight therefore adds roughly 190–330 t: the 120 t ship plus 70–200 t of payload, or propellant for tanker flights.
• Non-Starship launches (Falcon 9, China, New Glenn and others) are likely to total about 6,000–12,000 t by 2030.
• Getting past 100,000 t in one year therefore needs roughly 300–450 orbital Starship flights in a single calendar year.

Which years can count. A calendar year has to be finished before the publication that resolves the question. So only 2027–2030 can qualify. Their results would appear in GCAT updates in the first quarter of the following year. 2031 cannot count, because it will not be complete by 2031-12-31.

How I weighed it

Reference class. I compared Starship with past launch-rate ramps.

• Falcon 9 grew about 1.4–2 times a year: 31 flights in 2021, 61 in 2022, 96 in 2023, 134 in 2024, 165 in 2025.
• No vehicle has ever exceeded 165 orbital launches in a year.
• Starship has flown about 4–5 times in 2026. Polymarket has fewer than 5 at 51.5% and 5–6 at 38.5%.
• It has not yet caught or reflown a ship, or transferred propellant between ships.

Company targets have repeatedly been cut.

• Musk's 2028/2029/2030 targets are 1,000, 3,000 and 8,000 flights.
• On Aug 4 he said "at least one flight a day" within a year. By late September that had become a "weekly or twice-weekly" pace in 2027, which analyst Tim Farrar says is at best an end‑2027 goal.
• His one-year-ahead Starship flight-count predictions have usually come true at about 20% (e.g. roughly 25 forecast for 2025, 5 flown).

Other forecasts.

• Goldman Sachs and JPMorgan, both underwriters of the SpaceX share offering, model about 2,750 launches in 2030 and more than 7,000 through 2031.
• Morningstar gives 43% to SpaceX abandoning orbital data centres and only 7% to its "moonshot" case of about 59,000 satellites by the mid‑2030s.
• An FAA bull-case forecast, quoted by an analyst, has 385 US launches plus reentries in 2030.

Hard limits.

• Currently approved Starship launches total about 145 a year: Starbase 25, launch pad LC‑39A 44, and pad complex SLC‑37 about 76.
• The FAA's proposed environmental-waiver rule could raise those limits but faces litigation.
• Starbase Louisiana targets 2029, and the KeepTrack analysis below sees 2030–31 as more likely.
• Other bottlenecks: propellant plants, routine ship reuse, and building more than 10,000 large satellites a year (Starmind satellites weigh up to 4 t each).

Model. I simulated the 2030 Starship flight count as lognormal with a median of about 100 (80% range roughly 30–320). I assumed annual growth of about 1.9 times, a per-flight mass of 190–330 t, and 6,000–12,000 t from other launchers. The chance of crossing 100,000 t comes out at about 0.3% for 2028, about 2% by 2029, and about 7–9% by 2030. Using a median of 150 would raise the final figure to about 14%. I trimmed slightly for the risk that GCAT stops being maintained, since BryceTech's payload-only figures would be much harder to push past the threshold.

What I'm assuming in the timing. Almost all of the chance sits in the January–March 2030 window (2029 data) and January–March 2031 (2030 data). I added a small allowance for catalogue lags spilling into the following quarters, and a token floor for resolution errors.

The case against my estimate

Higher. SpaceX has about $86B from its share offering, its own demand engine in Starmind and Starlink V3, and a deregulating government. Falcon 9 shows it can scale operations fast. If a ship catch on Flight 15 leads to reflights in early 2027 and Florida pads open that year, 500 or more flights by 2029–30 is conceivable, and bank models assume far more than that. This is why I am not below about 5%.

Lower. Every Starship timeline so far has slipped by years. The threshold also requires regulation, propellant supply, launch pads and satellite production all to scale together within four years.

Sources

• GCAT column descriptions and data (Release 1.8.8; data update 2026‑10‑11): https://planet4589.org/space/gcat/web/cat/cols.html
• Ars Technica on Flight 14 (2026‑09‑28): https://arstechnica.com/space/2026/09/starships-first-orbital-launch-gives-lift-to-spacexs-next-gen-starlinks/
• Space.com on the SpaceX earnings call (2026‑08‑04)
• TradingView/Stocktwits on the cut to the cadence forecast (about 2026‑10‑01)
• Teslarati on Flight 15 and Starmind (October 2026)
• KeepTrack on Starbase Louisiana and launch caps: https://keeptrack.space/deep-dive/starbase-louisiana
• Advanced Television on Musk's 1,000/3,000/8,000 targets (2026‑06‑09)
• Faster Please on the Goldman and JPMorgan models (2026‑07‑07)
• The Deep Dive on Morningstar's scenarios (2026‑06‑08)
• SpaceX EU prospectus (2026‑06‑05)
• Polymarket gamma API (2026‑10‑13)

Question. By the last day of each quarter, has the total mass launched to orbit worldwide exceeded 100,000 tonnes in a calendar year?

Definitions.

• Tonne: 1,000 kilograms; the threshold is strictly greater than 100,000,000 kg.
• Total mass launched to orbit: mass delivered from Earth to an Earth orbit, measured at initial orbital insertion. Include payloads, spacecraft and reusable orbital vehicles, cargo, propellant remaining at insertion, rocket stages, and other hardware that reaches orbit. Exclude propellant consumed before insertion and hardware that does not reach orbit. This is total inserted mass, not merely commercially useful payload or advertised launch capacity.
• To orbit: successful insertion into Earth orbit, including parking, transfer, and deliberately short-lived orbits. Use Jonathan McDowell's GCAT classification to distinguish orbital from suborbital or failed-to-orbit missions. A subsequent departure for the Moon or another destination does not disqualify mass that first reached Earth orbit. Direct escape missions without Earth-orbit insertion do not count.
• Worldwide / calendar year: all Earth launch sites, operators, and countries combined, for launches dated January 1 through December 31 UTC. Assign mass to the Earth launch date, not the date a carried satellite is subsequently deployed. The calendar year must be complete by the resolving publication date.
• Counting: count each physical mass once per Earth-to-orbit trip. A vehicle that returns to Earth and is relaunched may count again; transfers between spacecraft, later deployments, fragmentation, and orbital maneuvers do not constitute additional launched mass. Include successful orbital delivery on a mission with other failures, but only the mass actually delivered.

Resolves YES if a qualifying source publishes an annual total above the threshold under these definitions, or its published data establish such a total by reproducible addition. A documented subtotal above the threshold suffices even if some masses are unknown. Published numerical mass estimates count; where a source gives only a range, its lower bound must exceed the threshold. A rounded value of exactly 100,000 tonnes does not suffice.

For a calculation from GCAT, use its object catalogs and launch lists, joining objects to their Earth launches and summing the Mass field, defined as mass at initial orbital insertion excluding attached objects. Count attached objects separately where their mass is excluded from the parent. Do not also add TotMass, repeated event records, or fragments whose mass was already counted in their parent. Use TotMass instead of constituent masses only for a documented non-overlapping assembly. Unknown masses contribute no inferred amount.

Resolves NO / Exclusions. Resolve NO at a horizon without qualifying evidence above the threshold. Exclude launch forecasts, annualized quarterly rates, maximum vehicle capacities, launch-pad gross mass, suborbital tests, failed orbital deliveries, and material produced or mined in space. A payload-only statistic below the threshold does not establish a total-mass result; a qualifying payload-only subtotal above it does establish YES. No attribution to AI is required.

Resolving source. Jonathan McDowell's General Catalog of Artificial Space Objects, its dated data releases and accompanying analyses in Jonathan's Space Report; or BryceTech's quarterly Global Space Activity and annual Orbital Launches Year in Review reports on brycetech.com. GCAT's Object Catalog Column Descriptions specify the mass fields. These are the permitted sources, rather than the developer-focused Public confirmation convention A2. Either source can establish YES if its figure satisfies the definitions; do not add their overlapping estimates together. A dynamic database requires a dated release or archived snapshot supporting the calculation. Apply the Resolution window, Observation period, and Resolution date convention A3. Later revisions may establish YES on their publication date; once YES, always YES.