Jul-2007

FCC gas concentration unit 
stripper revamp

Unit charge rate and reactor conversion were increased by the elimination of stripper column flooding. Chemical-grade propylene specifications were met

Dave Langdon, Total Lindsey Oil Refinery
Scott Golden & Edward Hartman, Process Consulting Services

Article Summary

The FCC gas concentration unit (GCU) stripper column at the Total Lindsey Oil Refinery (LOR) was revamped in 2005 to eliminate flooding, consistently maintain C2s in GCU mixed C3s at 5000 ppmwt or less (ie, <5000 ppmwt C2s in C3 mixture) and maximise C3 recovery (Figure 1). Flooding in the stripper column limited the FCC unit’s (FCCU) charge rate and reactor conversion prior to the revamp, requiring a bypass from the main column’s overhead receiver to the bottom of the stripper to manage flooding. C2s excursions in the GCU mixed C3s exceeding 20 000 ppmwt were also noted. Since LOR produces chemical-grade propylene (C3=) requiring C3= that exceeds 95%/ purity with total C2s and ethylene specifications of 400 and 50 ppmwt respectively, these 20 000+ ppmwt C2 excursions were unacceptable, as they overloaded the C2 stripping capacity of the two downstream propane/propylene (PP) splitting units. FCCU conversion had to be reduced to the stripper column’s hydraulic limit to meet the C3= product specifications. Following the revamp, stripper column flooding was eliminated, total C2s in the GCU mixed C3s were reduced to an average of 3000 ppmwt, and the FCCU’s reactor charge rate and conversion has increased to record levels. The stripper column’s feed rate is 33% higher than before the revamp, with no further flooding, and is no longer the FCCU’s constraint. 

Pre-revamp operation
Prior to the revamp, the stripper column limited the FCCU’s charge and conversion for several years. During the previous turnaround, the column internals were revamped to increase capacity, but the stripper did not meet its design objectives. Following the 2000 revamp, the reactor’s outlet temperature (ROT) had to be reduced 15°F (8°C) to lower the gasoline and LPG yield. Additionally, a wild naphtha product bypass line was installed from the main column’s overhead receiver to the base of the stripper column to reduce the stripper’s feed rate (Figure 2). Even with these changes, the stripper column limited the FCCU’s charge rate and conversion.

While the process flow scheme in Figure 1 is a generic depiction (eg, only one C3 splitter is depicted), the Lindsey facility actually has two PP splitter units (noted as C3 splitter in Figure 1), each with a different arrangement. They both serve the same purpose, but one utilises three towers: one tower operates as a separate deethaniser, followed by a two-shell PP splitter, essentially one for stripping and one for rectification. The other PP splitter unit achieves the same results in just one tower, with a C2 stripping section provided at the top of the tower and the C3= drawn as a side product.

Absorber/stripper systems
Primary absorber/stripper systems are the heart of the FCC GCU. These columns and the high-pressure (HP) receiver control C3 recovery and C2 rejection. Ideally, the stripper bottoms product’s C2 content is controlled at the maximum required to meet the GCU mixed C3s total allowable C2s of the previously noted 5000 ppmwt, because this maximises C3 recovery. Both columns are prone to flooding, with the stripper being the most problematic industry-wide. Primary absorbers operate at extremely high internal liquid rates, have moderate foaming tendencies, almost always flood due to inadequate downcomer top area or downcomer backup, and often flood at the intercooler draw or returns because of inadvertent restrictions. The stripper column also operates at extremely high liquid rates, with flooding initiated by either downcomer choke or backup flood mechanisms.

Rarely is jet flooding a problem in either the primary absorber or the stripper column. Moreover, the stripper can entrap water because the top temperature is too cold for all the water to go overhead, and the bottom temperature is too hot to allow the water to leave the bottom. This forces a water phase in the top and middle sections of the column, leading to severe flooding. When the stripper floods, it is not uncommon to rapidly fill the HP receiver, because the column feed rate is higher than the tray capacity. When the stripper floods, the FCC’s charge rate must be lowered and the reboiler’s heat input drastically reduced. Unfortunately, large reductions in the stripper heat input can cause major excursions of residual C2s in the mixed C3s product exiting the GCU. In this case, total C2s increased to 20 000 ppmwt or higher.

The stripper experienced severe flooding following the 2000 revamp. In an effort to mitigate this, LOR installed a bypass to route some of the wild naphtha around the primary absorber/stripper columns, thereby reducing the stripper’s feed rate. The wild naphtha maximum bypass was about 40 m3/hr out of a total stripper feed of 280 m3/hr. Therefore, the stripper’s feed rate unloaded by up to 15%. However, as a liquid proportion of the main liquid to the primary absorber, the reduction was 25% or more, which reduced the primary absorber’s internal liquid/vapour (L/V) ratio, thus reducing C3 recovery. Furthermore, the wild naphtha contained more than 2000 ppmwt C2s and high H2S, forcing the stripper column to overstrip its feed to meet the mixed C3s product’s C2s specifications. While this improved the overall performance of the FCC, the C2s excursions still occurred.

Primary absorber/stripper system revamp
LOR’s 2005 revamp installed new internals in the primary absorber/stripper columns, as well as a riser on the stripper feed draw nozzle in the HP receiver to improve oil water separation. Primary absorber modifications included new trays with a larger downcomer top area and modifications to the intercooler draws to eliminate restrictions. These changes have allowed the FCC’s feed rate and conversion to be increased without having to operate the wild naphtha bypass. Since startup, the unit has been tested at a maximum charge rate and conversion without any indications of either columns flooding. However, further charge rate increases have pushed the absorber/stripper system to a cooling limit. Thus, during summer operation, the GCU’s offgas C3+ content has increased from 8.0 to 12 mol%. Since this offgas is fed to a cryogenic gas plant where most of the C3+ is recovered, the impact of reduced GCU C3= recovery on the overall C3= product yield is less discernible. This said, the GCU’s operating objective is still to try to minimise C3+ from the sponge absorber, because C3= recovery efficiency on the cryogenic unit is only 70% maximum. So directionally more C3= leaving the sponge absorber will result in more C3= loss to fuel, with significant economic penalty. Given the overall FCC economics, the C3= losses do not impact the feed rate/ROT policy despite the GCU’s efforts to drive the losses from 12% back down to 8.0%.