In connection with the question on Starlink constellation potential contribution to a run-away phenomenon, also known as “Kessler syndrome”, it would be of interest to know how well actual collision statistics correlate with Kessler’s model.

In his 1978 paper Collision between artificial satellites: the creation of a debris belt, Kessler predicted (using the list of objects cataloged and space activity growth in the 70s) that the first random collision in space would occur between 1989 and 1997. He then concluded that:

Thus, unless significant changes are made in the method of placing objects into space, fragments from inter-collisions will become a source of additional space debris by the year 2000, perhaps much earlier.

This means that, up to a critical number of artificial satellites (and fragment objects), even if we were to stop launching additional ones, the number of objects in orbit would still mechanically grow, as a result of inter-collisions, leading to a "run-away" situation, i.e. an uncontrolled chain reaction.

Compared to the database used by Kessler in 1978, the number of objects larger than 10cm is now (Aug 2021) above 34 thousands, whereas there are almost a million of them between 1-10 cm size (hence difficult to track with current technology, yet can be harmful), according to ESA

This year, the Chinese Yunhai 1-02 satellite collided with a small debris fragment. Another well-documented event is the 2009 collision between an active Iridium satellite and a derelict Russian one.

  • Question: When did the first random collision between man-made objects (resulting in fragmentations) occur?

2 Answers 2


I think that there must be many such collisions that occurred without even the satellite operators noticing, and certainly not reporting them. There were also intentional collisions with anti-satellite tests even back in the 1960's.

The earliest I could find was a 1996 collision with Cerise. The Orbital Debris Quarterly News (a NASA publication) called it

...the first time that two objects in the U.S. satellite catalog are known to have accidentally run into one another.

I will also give an honorable mention to the 1983 mission of STS-7. Considering that to be an operational satellite at the time, it returned to earth with damage to a window from orbital debris.

  • $\begingroup$ The 1983 STS-7 mission did not result in lost of control and/or fragmentation (?), so Ok for "honorable mention". Hence, the earliest known event of interest could be the 1996 Cerise event, or before that (which then confirms Kessler's prediction). May be you can address the subsidiary question: has a chain reaction started? $\endgroup$
    – Ng Ph
    Aug 27, 2021 at 9:56
  • $\begingroup$ In all fairness, I have accepted your answer, although it does not address the subsidiary question. $\endgroup$
    – Ng Ph
    Aug 27, 2021 at 20:55
  • $\begingroup$ I don't have access to the original paper. I'm not sure what limits there are on "run-away". Does the first damaging impact from previous damage count (that has happened). Is the rate of damage going up (certainly). I don't know if either of those count (or if something in particular would count) as "initiation of the run-away mechanism" $\endgroup$
    – BowlOfRed
    Aug 27, 2021 at 21:09
  • 1
    $\begingroup$ Wasn't the second Soviet space station Salyut 2 (actually Almaz/OPS-1) hit by a cloud of debris from its own third stage detonating and subsequently destroyed in 1973? $\endgroup$ Aug 28, 2021 at 1:40
  • $\begingroup$ @BowlOfRed, never mind (I removed the subsidiary question). You may find Kessler's 2010 paper a better read to understand the complex mechanism leading to a "run-away". Loosely speaking, it is a situation where we can no longer stop the exponential growth of debris. $\endgroup$
    – Ng Ph
    Aug 30, 2021 at 21:01

This answer is now obsolete as question has been clarified.

First collision between an operational satellite and "another object in space"?

all of them, most likely!

For starters, in the period from February 1 1958 to February 12 1958, Explorer 1 detected 145 micrometeorite impacts. Micrometeorite detector experiment on Explorer-1

I think the OP might want to rephrase that "another object" stipulation with some parameters for size and origin. And provide detail on what will qualify as proof for the artificial origin of an object, if OP wishes to make that distinction.

  • $\begingroup$ Point taken (see edits). $\endgroup$
    – Ng Ph
    Aug 27, 2021 at 8:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.