8 replies to this topic

The vacuum unit feed is residue from atmospheric crude distillation and has typical setup of wash bed, hvgo wash oil. May I know what are the common factors that affect overflash/slop wax rates (while operating conditions are maintained) and what are the means of controlling overflash/slop wax rates effectively? Also, can feeding different kinds of crude to the crude distillation unit drastically affect overflash in the vacuum unit?

[Zauberberg]

Sardonyx,

1. Yes, the type of feed (long residue) affects wash oil flow requirements, it is straightforward dependence. More volatile feed increases vapor superficial velocity of flash zone vapors and promotes higher entrainment rates of black oil product (vacuum residue).

2. Usually, there is a total draw-off collector tray beneath wash oil bed (packing) and liquid stream (slop wax) is recycled back to the furnace inlet, and commingled with fresh feed. This is the most economical option because it yields in maximum HVGO recovery. Wash oil flow should be adjusted to meet HVGO product quality targets: metal content (especially if this stream is routed to Cat Cracker Unit - severe catalyst poisoning may occur), colour, Conradson carbon, ASTM D1160 distillation end point etc. Wash oil flow should be flexible in order not to loose distillate yield without reason, but there is a minimum value which ensures complete wetting of structured packing bed. Below this flowrate, coking inside the bed will start and once when this happens you'll see that even excessive flowrates of wash oil cannot restore your HVGO quality.

More information you can find on Process Consulting Services, Graham-MFG and Karl Kolmetz websites. Additionally, ensure that Norm Lieberman's and Henry Kister's books are in your private library.

Regards,

[djack77494]

Like all design values, what you chose will be a compromise. In my experience, a commonly used target is to obtain 0.15 to 0.20 USgpm/ft^2 of packing (grid) cross sectional area at the bottom of the packing. (This value is MUCH higher at the top of the packed section.)

[Zauberberg]

The purpose of washing flash zone vapors is to clean the distillate from entrained residue, with as little reflux (wash oil) as possible and with as low pressure drop as possible. And you must avoid, at any cost, the coking of wash oil bed; otherwise VDU has to be shutdown or heater outlet temperature needs to be reduced, in order to produce HVGO with acceptable quality. In any way, you will reduce HVGO yield.

Usually, wash oil bed is designed to have at least 0.3 gpm wash oil flow per square feet of top packing surface. However, if wash bed is unreasonably high (someone may think that increasing bed depth will provide additional protection against entrainment - don't do that!) the coking will start to occur because complete wash oil will re-evaporate while travelling through the packing.

[Zauberberg]

Here are the links to some articles that might be of interest to you:

http://www.revamps.c...cuments/167.pdf
http://kolmetz.com/p...ervicerev99.pdf
http://www.graham-mf...wnloads/210.pdf
http://home.earthlin...nti/GVMFeed.pdf
http://home.earthlin...anti/Refine.pdf
http://kolmetz.com/pdf/odcdrev65.pdf

Hope this will provide you some additional answers.
Best regards,

Wow thanks guys, you are a tremendous help.

I am a newbie in this field and have problems with sometimes wildly fluctuating overflash rates (eg the rate can change from 0.2gpm/ft2 to 0.4 for sustained periods of time or drop to as low as 0.05 which may cause the column to shut down prematurely). Wash oil rates and feed temperature have been adjusted to little effect and the bed and nozzles were cleaned not long ago. The reasons for difference in overflash rates are largely unknown though it seems to have some loose relation to different types of crude used (though crudes used have similar density, cut points etc).

Have anyone encountered such problems? What may cause inability to generate overflash?

[Zauberberg]

Sardonyx,

How do you measure the overflash? Do not mix it with slop wax recycle, as these two are not the same.

How do you calculate gpm/ft2 of wash oil leaving the wash section?

By overflash i mean wash oil + entrainment and other contaminants that are coming down from the bottom of the packing and exiting from the tray below (I think some people call it slop wax?).

The volumetric flux is calculated from the above overflash flow rate and the wash bed cross sectional area.

[Zauberberg]

From your statement I understand you are measuring actual slop wax flowrate and divide this number with wash zone cross-sectional area, in order to calculate the value of overflash. This is not correct. The actual overflash is about 60-70% of the value calculated this way (it can be quite different in cases of excessive or extremely insufficient wash oil rates). There is usually one volume of entrained liquid for each two volumes of residual wash oil and this mixture called "slop wax" is recycled back to heater inlet line.

The best way to calculate actual oveflash is to determine heavy metals content in HVGO, Slop Wax and Vacuum Residue, and metals balance will give you relative flowrates. This is way straightforward. Another option is to perform heat balance calculations around wash section, if flow measurements and temperature transmitters are correct.

Nevertheless, I still have some questions:

1. You didn't explain how this "wash oil fluctuation" - as you call it - affects HVGO quality? Any changes in colour, draw-off temperature, metals and concarbon values?

2. Have you noticed any changes in wash section pressure drop during these upsets?

3. What kind of vacuum tower is there in your plant? Trayed or packed, with slop wax total draw-off (chimney tray), do you quench tower bottoms etc. Can you give some more information?

Be sure to obtain at least two sets of HVGO, Residue and Slop Wax samples and send them to the laboratory for heavy metals analyses (Ni, V, Fe). And you do this for each type of feed you are processing.

Best of luck,