I think I have explained this before. In Sedimetentology we call this "Early Marine Diagenesis"
1. The local water in the SB develops a low pH from activity. This causes the dissolution. As dissolution continues the local pH, Alk and Ca++ begin to rise and reaches saturation, causing the precipitation of some CaCO3 around the grains "welding" them together.
2. Fresh new carbonate sands, like in a new tank, are not in equilibrium with the water. This causes the direct precipitation of calcite and often Hi-Magneisum Calcite on the the sand grains surfaces " welding " them together. There are many paper written on this and it is this that is more than likely what is happing in new tanks with new sand bottoms of carbonate sands.
3. Excessive Ca++ Alk and high pH can also cause this, as the sand favors more to precipitation in the SB, which will be even greater if the temp is a little higher there.
The above three produce a cement like clumping structure, where the grains can not be separated without hammer in hand.
4. Certain types of high SB bacterial activity produce organic "slimes / glues" that stick the sand grains together. In this case of SB clumping the grains can be pulled apart. The are more like a sticky cluey mess.
Forgot
Kalk with vinegar or other types of carbon dosing could make this worse where there may be more than one thing going on.