Without much more information, I would imagine that a few things would be at play:
1. Speed. Going too fast, this would cause more water to try and go up the scoop than it can handle. The result would be the excess water finding another route, spray, etc.. because water is incompressible.
2. Depth of the scoop, relative to the water level. If the scoop was too shallow, I imagine it would tend to shear the water surface, without much going up to the tender. Kind of like when you row a boat, and 'miss' the water. If it was too deep, a lot of force resisting the movement of the scoop through the water resulting in an instantaneous braking effect.
We studied this specific application in freshman physics. It seems to me that there were several approaches to how to introduce the pan water to the tender. I recall that they did have to open the regular water hatch when filling with the track pan just to reduce the chance the tender would blow up once it reached the full level.
One thing that wasn't discussed is the effect of the cinders from passing trains being thrown in to the pans. Since coal is less dense than water, it tends to float. That means the scoop is picking up a lot of particles that end up in the boiler. Maybe it was negligable.
