what's the fluid temperature?

No definite number there but my colleague suspects that, "they’re not far above boiling." I think he's guessing that because of a pressure tank that's in the system.

"Empty your mind, be formless, shapeless — like water. Now you put water in a cup, it becomes the cup; You put water into a bottle it becomes the bottle; You put it in a teapot it becomes the teapot. Now water can flow or it can crash. Be water, my friend." -Bruce Lee

The image looks like an incompatible material issue.

Tungsten carbide in not compatible with: TUNGSTEN CARBIDE reacts violently with F2, ClF3, IF5, PbO2, NO2 and N2O (oxidants). It is readily attacked by a nitric acid-hydrofluoric acid mixture. May ignite in cold fluorine.

did they forget to turn on the seal flush line? Or are feeding water close to boiling or even supercritical to the seals?

The picture shows fractures directly comparable to what you get when something metallic get a mechanical induced shock. A chemical reaction would in my opinion show other types of damage on the surface in addition. Searching I found an article on how 'sturdy and solid' wedding rings of tungsten carbide breaks when they hit concrete.

My suspicion is then some kind of mechanical shock. Could be caused (guessing) by sudden thermal shock (different crimping or enlargement), different sudden speed changes (hanging) by start/stop, different kind of shocks in fluid (start, stop and/or waterhammer), drying up giving start-up disturbances etc.