Fouling and Plugging Cause High Reactor Pressure Drop and Premature Shutdown in Hydroprocessing Units

By Joseph J. Kociscin

Carmagen was retained by a US-based Refiner to conduct a "Cold Eyes Review" of their Cat Feed Hydrotreater (CFHT) reactor fouling and pressure drop problems.  High reactor Delta Pressure (DP) buildup required several premature shutdowns.  In any large refinery, a premature shutdown to clean a unit or replace/skim reactor catalyst can cost hundreds of thousands of dollars per day in lost revenues and expenses.

This unit processes about 50% coker products in the feed.  It was designed to operate at 45kB/D, and revamped several years ago to process 70kB/D without adding preheat exchangers or paralleling the two-series reactors. Prior to the most recent cycle, DP was limiting across the first bed; therefore the refinery removed the inter-bed quench and re-distributor assembly from the first reactor R1, but the unit DP problems persisted.  There were attendant flow and temperature maldistribution problems in the reactor, with excessive DP in R2 and a hot spot in R1. Carmagen recommended alternative solutions and general recommendations to eliminate those problems. They included modifying the bed grading scheme with fewer gradings and without the use of “virtual scale” baskets, reinstalling the first reactor inter-bed assembly and including proprietary vapor and liquid bypass tubes for the top bed, reconfiguring the reactors for parallel operation and eliminating the multiple grading scheme, inspecting and leak testing the existing distributor trays, and evaluating alternate disposition of coker naphtha and light coker distillate feeds. Implementing appropriate items above would be necessary to achieve a target 3-year run.

The major fouling precursors are entrained small solid particles in the filtered feed (smaller than 25 microns), plus dissolved chemical components that react to form smaller and larger insoluble solid reaction products from iron naphthenates, diolefins, and other reactive cyclic and aromatics with alkene substituents.  Reactors loaded with the equivalent of 1/16 inch cylindrical- or shaped-catalyst extrudates will pass particles smaller than 150-200  microns, depending on the catalyst compaction, shape and length distribution.  However, when the concentration of <25 microns solids gets too high, they will also accumulate and cause DP in the catalyst bed, particularly in pockets of higher density or lower void fraction than the bulk catalyst average.

In addition to the fouling and plugging work, we were asked to conduct a technical comparison of new distributor trays being considered to replace the existing bubble cap trays. During the last 10 years, many hundreds of commercial sieve distributor trays, tube and chimney distribution trays were replaced with the new generation distributor trays, which resulted in activity credits from 10 to 35%. Similar significant credits were also achieved when old leaky trays were simply repaired. Vendors and licensors of the new generation vapor/liquid distributor trays take advantage of the kinetic energy of the vapor (or treat gas) flowing through the tray tubes or nozzles. Various devices are used (e.g. the spiral mixing type in the VENDOR A design, the entrainment or liquid lifting type of “U-Tubes” in the VENDOR B design, and the downcomer tubes with holes and slots to generate a mist of small droplets that sprays out of the nozzle as in the VENDOR C and VENDOR D designs, and the pigtail spray nozzle tray arrangement in a second VENDOR C design). In my opinion these vapor-assisted, liquid distributors are superior to the simple multiple short and long tube designs, the sieve tray designs, and the reverse flow bubble cap design. However, the new and some existing reverse flow bubble cap designs, with a smaller pitch, seem to be doing quite well. In many cases, there does not appear to be any big difference among the new tray designs, but the cost difference could be significant.