(206e) Reverse Vapor Cross Flow Channeling in High-Capacity Trays: Diagnosis and Fixup | AIChE

(206e) Reverse Vapor Cross Flow Channeling in High-Capacity Trays: Diagnosis and Fixup

Authors 

O'Shea, N., Phillips 66 Whitegate Refinery
Cronin, D., Conoco Phillips Ireland Limited

This paper originates from a systematic troubleshooting effort that successfully diagnosed two complex independent problems which severely limited the capacity of an atmospheric crude distillation tower. The crude tower had been retrayed (by others) with modern high-capacity trays, but upon start-up achieved less than 80% of its previous capacity.

The troubleshooting identified two bottlenecks: premature flooding on some of the heavy naphtha-diesel fractionation trays, and a phenomenon which we termed “excursion,” in which sudden temperature reductions around the diesel draw, accompanied by distillate yield loss, occurred upon rate increase.

Our previous paper (Kister, H. Z., N. O’Shea, and D. Cronin, “Systematic Field Tests Turn Loss into Gain in High-Capacity Trays, Part 1: Excursion”, Proceedings of the Distillation Topical Conference of the AIChE National Spring Meeting, Houston, TX, April 2007) focused on diagnosing and fixup of  the excursion bottleneck. This paper focuses on identifying and fixup of the high-capacity trays bottleneck.

Multichordal gamma scans with quantitative analysis (Kister, H. Z., “Use Quantitative Gamma Scans to Troubleshoot Maldistribution on Trays”, Chem. Eng. Progr., February 2013) provided detailed mapping of the froth heights, froth densities, hydraulic heads, and entrainment profiles on the trays. These profiles showed conclusively that the trays were bottlenecked by a previously unknown mechanism which we termed “Reverse Vapor cross Flow Channeling (RVCFC)”. This phenomenon is closely related to vapor cross flow channeling (VCFC), but is induced by the forward push on high capacity trays rather than by the hydraulic gradient on the trays. Like VCFC, this phenomenon sets in when there is a significant channeling initiator while the friction pressure drop of the trays is too low to counter the channeling and when the ratio of flowpath length to tray spacing is high.

Based on this diagnosis, the tower was retrayed with the same type of tray but with a design that gave significantly larger friction pressure drop. The new trays also had two passes which halved the ratio of flowpath length to tray spacing. The modifications completely eliminated the bottleneck and the RVCFC, permitting higher-than-design capacities to be achieved.