The star Fomalhaut is about 25 light years away and twice as massive as our Sun. It is considered by astronomers to be a few hundred million years old. Images from the IRAS infrared satellite in the 1980s indicated that the star was surrounded by large amounts of dust. ESA's Herschel Space Observatory has provided an even higher resolution image of the star and its dust disk. It shows the star to be surrounded by hot dust and gas, but al so has a dusty belt of material at the outer fringes of the star system.
The belt is more than 100 times the distance between the Earth and the Sun makes it very cold, with a temperature of about -200 Celsius. About ½ of this material is water ice. Fomalhaut icy dust belt forms a relatively narrow off-centre ring with respect to Fomalhaut suggesting the presence of planets orbiting close to it or as suggested by the 2004-2006 Hubble imaging of a planet orbiting on the edge of the ring.
Based on the emitting, absorbing and scattering of the starlight by the disk the grains have been found to be as small grains a few thousandths of a millimeter across. However Hubble images had suggested larger particles. This indicates fluffy dust grains consisting of small particles stuck loosely together forming the larger ones. However these fluffy grains should have been blown away by the light pressure from Fomalhaut. The explanation proposed by astronomers is that continuous collisions and disintegration of larger asteroid-sized comets re-supplies the small particles.
However re-supplying the large amount of dust observed by Herschel would require the destruction of a whopping 2,000 1km across comets every day. Maintaining that many collisions per day requires trillions of comets to be orbiting inside the ring. That many commits would have a combined mass of over 100 Earths.
The point is that astronomers started with these two assumptions and there fore did not consider the other options. The result is a need for a minimum of a thousand of dally collisions of unobserved comets. That come to about 42 collisions an hour or about one collision every 87 second. This is an amazingly high collision rate. Even if this theory is true is presents a problem for planet formation theory since it shows that larger peaces material in these dust clouds tend to get broken up as opposed to forming new planets