We have a naphtha hydrotreater unit designed to process straight run and cracked naphtha. The unit has 2 flash drums (hot and cold separator drums). Both separators are operating at 2800 kPag. Both flash drum liquids are sent to the debutanizer tower.
We checked the thermodynamics of the ammonium chloride deposition in the unit. Based on our prediction using the calculated Kp and ammonium chloride sublimation curve, the deposition of ammonium chloride is most likely seen in the flash drum sections.
Circuit
Kp
Reactor to Flash Drum 1
9.12 E-06
Flash Drum 1 vapor to Flash Drum 2
2.42 E-05
Debutanizer feed
3.96 E-06
Debutanizer top tray
1.13 E-06
Note: Nitrogen and chloride present in the feed are 1 ppm and 0.2 ppm, respectively.
However, during our offline water washing, ammonium chloride deposition in the flash drum is not significant compared to the deposition on the top trays of the debutanizer. What is the possible explanation for this deviation of the actual deposition to the thermodynamic prediction? Also, besides concentration of nitrogen and chloride in feed and hydrogen gas, what are the other parameters affecting ammonium chloride deposition?