"Sand-type" coke morphology
Typically, we process high asphaltene, high-MCR vacuum residues in our Delayed Coker, that produce a shot-coke morphology.
But in one of our Delayed Coker Units we have recently changed feedstock quality, to light (low-MCR, low ashaltenes). This light VR is producing a transition coke (asphaltene / MCR ratio about 0,6) with low particle size (less than 1 mm), quite loose, that looks like "sand". This coke give us a lot of problems during coking cycle: high level alert that is not consistent with coke yield and drum filling (it seems as coke "floats" or is withdrawn with coking vapours) and more severe problems during decoking cycle: problems during cooling (high level measurement, it seems as coke "floats" in water), difficulties in cooling (more time required, it seems as coke release more heat that usually), plugging problems in drainage line and bed collapse during coke cutting.
Definitively, the coke bed formed is very loose, not compact and "mobile".
We suspect that it is due to: 1) Excessive velocity in the coke drum (due to higher gas production); 2) Light feedstock require more time in order to obtain a compact coke.
How could we improve the coke morphology to avoid this problem? Which should be the changes in operating conditions to avoid these problems?