UPDATE IN RESPONSE TO YOUR COMMENT
I apologise : as you suggest, there might be a lift force on the sphere if there is a shear flow in the fluid (see Discussion in 1st Link). However, this force is likely to be much smaller than the drag on the particle.
I do not see any reason for there to be a lift force since the particles do not appear to have an aerodynamic shape. Even if they did, the fact that they are free to rotate means that the 'angle of attack' could be reduced to zero, which would reduce lift to zero.
Possibly the particles could spin which would introduce a force due to the Magnus Effect (see video link). However, if the particles are quite small, viscous drag would quickly wipe out the rotation.
Your buoyancy term should use the difference in density between the particles and the fluid.
Drag is friction which opposes relative motion. Assuming the particles are denser than the fluid the drag force should probably be in the direction of fluid flow around the particles, which may not be upwards. This direction depends on position in relation to the step, and would be difficult to predict, even when the flow is 'laminar' rather than 'turbulent'. The particles themselves might also disrupt the pattern of fluid flow. If the particles are initially on the bottom of the step, they might be 'dragged' into the corner but probably not upwards - unless they are almost floating and/or there is a lot of turbulence.
Remember that your equation of motion is a vector equation : forces ma and Fd both have x components as well as y.