Deployment path
What you can promise before the shell is finished
Every other figure on this site describes a completed constellation. Nobody operates one for years, and those years are when coverage gets sold. Set a launch plan and see the service the fleet actually delivers at each milestone, against the number a specification sheet quotes for the same satellite count.
Deployment path
The first milestone
After launch 1, 8 satellites in 1 plane, the worst outage at 50° is 17.3 h. A specification sheet quoting 8 satellites would say 45.7 min, because it assumes 4 evenly spread planes that do not exist yet. That is 22.7x worse than the number on the sheet.
Service through the buildout
6 launches of 8 satellites into one plane each, reaching 48 satellites in 6 planes. At 50° latitude, min elevation 25°.
Milestone by milestone
| Launch | Satellites | Planes | As deployed | As specified | Service min/day | Difference |
|---|---|---|---|---|---|---|
| L1 | 8 | 1 of 4 | 17.3 h | 45.7 min | 132 | 22.7x worse |
| L2 | 16 | 2 of 4 | 5.3 h | 21.0 min | 263 | 15.3x worse |
| L3 | 24 | 3 of 6 | 85.7 min | 20.7 min | 394 | 4.1x worse |
| L4 | 32 | 4 of 8 | 8.3 min | 10.0 min | 526 | 1.2x better |
| L5 | 40 | 5 of 8 | 8.3 min | 6.3 min | 581 | 1.3x worse |
| L6 | 48 | 6 of 8 | 8.3 min | 5.7 min | 500 | 1.5x worse |
Planes shown as populated, against the number the automatic rule would use for that fleet size. Both columns are measured at 8 longitudes and 20 s steps.
What this assumes
- Planes are evenly spread at every milestone. An operator building toward a twelve-plane shell puts its first three launches into three well-separated slots if it can. One constrained by launch windows or rideshare manifests does worse than this shows.
- Satellites are evenly phased within a plane from the moment they arrive. A real launch deploys them clustered, and they spread over weeks or months. Early milestones are therefore optimistic even as drawn.
- Nothing fails and nothing is retired. No replacement launches, no early losses, no station keeping.
- The specified column is the automatic plane rule, which is exactly what the rest of this site, and most specification sheets, quote for a given satellite count. Full method.
Why the gap exists
A launch reaches one orbital plane. Satellites in a single plane share one ground track, so a site gets a burst of passes while the track sweeps over it and then nothing until the geometry returns. Adding satellites to that plane makes the burst denser without shortening the wait between bursts.
A specification sheet quotes the finished shell, where the same satellites are spread across many planes and the waits between them are short. Both numbers are true. They describe different constellations, and for the first few years of a programme only one of them exists.
On the plan loaded above, the widest gap falls at launch 1, with 8 satellites in orbit: 17.3 h as deployed against 45.7 min as specified, a factor of 22.7. The gap closes as the shell fills, which is the point: it is widest exactly when the programme is out selling.
Using it
Quote the as-deployed column with a date against it. If a customer needs a particular worst outage by a particular quarter, this says which launch delivers it, and the answer is often several launches later than the satellite count alone suggests. Where a milestone matters, check whether a different plane split at the same fleet size does better: the calculator will search every arrangement for you, and the sizing tool answers the same question from the other direction.
Where the as-deployed figure is better than the specified one, which does happen, the automatic rule has simply chosen a worse arrangement than the buildout happens to have. Coverage depends on how planes interleave, not only on how many satellites are up.