It would be slightly more accurate to have two displacements for the sides, one that only sets X = 0 and another that only sets Z = 0 and instead of a Fixed Support on the bottom, that would be a third displacement of Y = 0. These may not make much of a difference as you make the solid boundary further away.
Don't plot Total Deformation, plot Directional Deformation in the Y axis. I expect that 0.15 m displacement is downward, is that correct?
Try meshing just a soil box with no raft in the model. What is the deformation then? Use that as a baseline to compare with the additional deformation when the weight of the raft is included.
You could begin the simulation with the soil prestressed with the state of stress from the gravity load using the INISTATE command. Look it up in the Mechanical APDL section of ANSYS Help. You can also find several posts on this in the forum. That way, when the simulation starts, the soil is already in equilibrium with gravity, and the only deformation is that needed to support the new weight of the raft.
Another approach is to use the information from the soil-only model for the vertical deformation due to gravity. Make a 2-step simulation where you have a displacement on the raft that moves it the same distance that the soil moves in step 1, then in step 2, remove the displacement constraint and let the raft sink further into the soil due to its weight. Now you can look at the difference in Y deformation between step 1 and step 2.
Try doubling the size of the soil box. Does that reduce the deformation?