vacuum column revamps 2

[refilwe]

The vacuum column in the slo fractionation unit is giving us problems. Some of the HVGO goes with the vacuum bottoms.
How can i fix this problem....

Thank you

 

[epoisses]

Show us a list of the things you have tried so far and we promise to show you a list of as many things you could try.

 

[25362]

To refilwe,

I wouldn't mind to know what slo stands for.
Anyway, can you tell us what type of operation is involved, dry or wet ? Temperatures and pressures, what are they, and are they steady ? Do you have precondensers or is the tower directly connected to a main "booster" ejector ?

Generally, some of the options to recover more HVGO are:

[•] Unload the tower and the heater by improving
previous atmospheric GO removal.
[•] Reduce thermal cracking which affects vacuum.
[•] Reduce [Δ]P and flash zone pressure by
switching from pumparound trays to packing.
[•] Reduce number of wash trays; results as above.
[•] Improve vacuum generation.
[•] Increase the superheat of steam to the resid
stripping trays, and check the cooling effect of this
stripping; below 15 Celsius means the contact with steam
isn't OK. It may be due to tray flooding, corrosion, or
upset. Even a bottom level controller malfunction could
be the culprit.
[•] Don't over-wash the lower fractionation section.
[•] Seal weld HVGO and dark or slop VGO draw-off trays
to minimize leakage into the residue.
[•] Quench the bottoms to avoid thermal cracking.
[•] Circulate slop VGO to extinction.

The general ROT to verify whether there is too much HVGO (up to 540 Celsius) in the residue is to vacuum distill it in the lab. 5 to 10% is reasonably good; more than 30%, points to a poor performance. It even can markedly reduce the NPSHA of the bottom pump.

You may find some of the above steps aren't applicable or contradict each other.

Please comment.

 

[epoisses]

Wow 25362, you deserve a star
I couldn't, and not only because I don't have your expertise...

 

[refilwe]

Thank you for the responses.

The operating conditions are as follows
Pressure: Overhead -52KPa Ab
Flash Zone -51kPag

Temperature: LVGO REFLUX 40 DEGRESS CELCIUS
HVGO REFLUX 86
BOTTOMS 260
HVGO HOT RETURN 155
The problem is that the vacuum unit experiences a low TBP cut point.

So far i was looking into increasing the furnace temperature and or reducing the flash zone pressure.

Thank you

 

[25362]

Refilwe,

Could you refer to the following queries ?

1. The vacuum level is not impressive. Is the hydrocarbons'
partial pressure being reduced with steam ?
2. What are the temperatures at the flash zone and the top?
3. Again, what is the meaning of the acronym slo ?
4. Have you checked the unit for inward air leakages ?
5. How do you intend to improve the vacuum ?
6. Are you using velocity steam in the heater ?
7. How many wash trays do you have, and of what kind ?
8. What do you do with the black gas oil from washings ?
9. Does the unit have a top precondenser ?
10.Why do you express the absolute pressures at the top in
kPa abs., and in kPag at the flash zone ?

Epoisses, thanx for the star, the truth is my "experience" is reinforced by plenty of notes I collected over the years.

 

[EmmanuelTop]

Refilwe,

What is the design cut-point for HVGO and vacuum resid?
According to your tower pressure profile data, decent HVGO recovery should not be expected. Or, possibly, you are suffering from very intensive condensate back-up in your vacuum condensers. Do you have 1-, 2-, or 3-stage ejector system?
Also, which is very important, are these P,T data from the DCS or directly measured outside on the process equipment? What is the design pressure of tower top and FZ?

Before answering, you may want to look at some of these resources, they could help you:

http://www.graham-mfg.com/downloads/216.pdf

http://www.graham-mfg.com/downloads/220.pdf

http://www.graham-mfg.com/downloads/221.pdf

http://www.graham-mfg.com/downloads/232.pdf

 

[0707]

Maybe your lower packing is plugged with coke, which hinders the HVGO to go up, or you are having problems with HVGO reflux. Try to make a gamma scanning to check for mechanical hardware damages and liquid misdistribution on lower part of the column.

 

[katmar]

The overhead pressure cannot be -52 kPa abs because you cannot go below zero. If both pressures are actually gauge pressures then the overhead pressure is lower than the flash pressure. If the overhead pressure is actually 52 kPa abs it would be equivalent to a gauge pressure of -49 kPag (assuming local atmospheric pressure is 101 kPa abs). This makes the overhead pressure higher than the flash pressure.

I know nothing about the set up of this particular application, so I cannot comment on which scenario makes more sense. But when the data is inconsistent like this it makes me suspicious of all the data. It is very hard to trouble shoot when you don't have a clear picture.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com