What? Starlink eventually wants 40,000 satellites - with 5 year lifetimes that'll be 8,000 per year... there are numerous other planned constellations, so figure something like 15,000 LEO per year within a few decades.
15k satellites per year with a 5 year lifespan and a 1-year decay means you'll be stable at 75k satellites. 5 year lifespan with a 3-year decay means you'll eventually be stable at over 100k satellites in orbit with 30,000 of them 'dead' and decaying. That's a massive difference in collision risk.
Is it? It's a big difference in absolute numbers but that doesn't mean it's a meaningful difference in Risk.
If both numbers are quite small relative to the level of concern, the difference can still be irrelevant.
My point is that it takes more than just looking at the number of satilites to understand risk. You need to do the work to show how the collision chance compares to what an acceptable limit might be. Both cases could be very acceptable (or both cases could be unacceptable).
The rule is 5 years. You've given 2 examples under 5 years. If everything is going to re-enter within 5 years without boosting then your point is moot, because even in your "massively more collision risky" example, it still abides by the rules.
If something was going to deorbit within 1 minute, or was going to deorbit within 4 years, 6 months... It doesn't matter to the creator, because they don't have to change their design at all to meet the rules.
If something was going to deorbit within 8 years (because they previously had a 25 year allowed limit), they now have to rework the design.
There's plenty of room for debate about if 5 years is adequate, but as it stands, _most_ things (under 500km) will naturally deorbit within the legal time frame anyway even without special consideration.
Just responding to a comment that said "If deorbiting in years is sufficient, the difference in time is not relevant" where it's obviously relevant.. this doesn't seem remotely controversial.
You're arguing a different thing. The topic was satellite design for satellites under 450km.
If deorbiting in years is sufficient, the difference in 1 minute vs 4 years is NOT relevant -> to a satellite builder worried about the law.
If everything deorbits within 5 years, the only way for more things to accumulate is to launch things faster. But that's a separate discussion.
If everything launched today is deorbited within 5 years, then in 5 years, all satellites will be new satellites launched after today.
If everything launched today is deorbited within 5 months, then in 5 years, all satellites will be new satellites launched after today.
Deorbit speed under a threshold has no bearing on accumulation beyond that threshold.
If SpaceX launches a trillion Starlink satellites, and they all deorbit within a year, then yes. it's going to be a very crowded year, and we'll have to drastically rethink how much stuff we have in LEO, but at the same time SpaceX would not be in violation of the 5 year deorbit window, so the issue is about how much stuff we're sending up, and not how fast it de-orbits.
"Amount of junk below 450km, total" and "Amount of junk below 450km, that hasn't deorbited after 5 years" are very different things. You're making points about total, while the original point was about deorbit speed.
15k satellites per year with a 5 year lifespan and a 1-year decay means you'll be stable at 75k satellites. 5 year lifespan with a 3-year decay means you'll eventually be stable at over 100k satellites in orbit with 30,000 of them 'dead' and decaying. That's a massive difference in collision risk.