Prospectus

Return to Table of Contents Volunteer to help write Position Paper

DSP-17 Spray Nozzle Entrainment from Low Surface Area Y and X Style Structured Packings, Grid and Random Packing at Deep Vacuum Expected Benefit toMembers: To quantify/compare the amount of spray nozzle liquid re-entrainment from the surface of a large crimped size sheet metal structured packing (both “Y” and “X” styles), grid packing, and random packing. This research would help fractionation specialists and designers specify the appropriate packing type at the top of a packed bed to minimize the loss of wash oil wetting (due to spray droplet re-entrainment) in low liquid rate wash zone sections in critical deep vacuum fractionator services. Present Situation and Proposed Research: Spray nozzle performance has been characterized at FRI in 1966, 1982, 1985 and 1997. The 1985 tests employed 1 and 7 nozzles from Spraying Systems to characterize spray distribution quality on the performance of 1-inch Pall Rings. HETP’s with the 7-spray nozzle system were better than with the single spray nozzle. In 1997, heat transfer studies were performed using a short bed of Intalox 4T structured packing and three single wide-angle spray nozzles. Test systems included C6/C7 at 24 psia totally condensing system and partially condensing C6/C7 with methane gas. Heat transfer was almost instantaneous for both systems. The tests in the 4 ft. LP section of the FRI unit in Stillwater, OK will simulate wash oil entrainment in which the entrained drop size distributions can be measured using the FRI/OSU owned Phase Doppler Interferometer (PDI). The pressure drop across the spray nozzles and the liquid rate would be metered for vapor rates (Cs-Factors) from 0.1 to 0.4 ft./s. Proposed Internals and Test System: Traditional Research Idea

Estimated Unit Time: o/p xylene system at 65-75 mm Hg. Estimated Additional Costs (Beyond Unit Time): One operating week for each internal, total unit time 10 weeks.

Lechler 423.148 or 423.208 spray nozzle as the liquid distributor; Flexipac ® 3YS or Mellpak™ 125 Y (smooth surface) or equivalent, Flexipac ® 3XS or Mellpak™ 125 X (smooth surface) or equivalent, Flexigrid ® #3 or F-Grid 3 or equivalent, 3rd or 4th generation random packing.

Background and Discussion: Optimizing the wash oil flow rate is key to preventing coking in a vacuum pipestill wash bed. To prevent coking in a wash bed, the desired true overflash (not including entrained residuum) is 0.1 gpm/ft2 (minimum) leaving the bottom of the packed bed. However, uncertainty about the amount of entrained residuum in the overflash, along with the notoriously poor reliability of overflash flowmeters, can make it extremely difficult to determine whether the 0.1 gpm/ft2 target is reached. Insufficient liquid to remove entrained residue from the wash bed packing leads to the formation of coke. Accelerated wash bed coking can be attributed to the loss of wash oil due to vaporization or re-entrainment from the top of the packed bed.

PROSPECTUS

Page | 13