3 replies to this topic

[nick_tri]

Goodday all,
I've been tryin to figure out how to design a Feed Surge Drum. I took note of DJack77494 comments on a similar question like this back in Nov 2005. However since I require a detailed design of this drum and I have little knowledge on how to approach it, I've been looking everywhere for data. I tried to size it as a Fractionator Reflux Drum (from McKetta) but I believe I can't use my inlet streams as Reflux and Overhead streams. So this seemed to be no help.

In "Rules of Thumbs for Chemical Engineers" I saw they have calculations for Vapor/Liquid Vertical Drum or Horizontal Drum. I don't know if this approach is a good way to go. I did the Vertical Drum calculation but was not quite sure where I was going with this since i assumed a (design time to fill of 10 min.) and using my liquid volumetric flowrate I found the volume to be 1.9 m3. The problem with this was that I have a very small amount of vapour in the form of Acetone goin into the drum and not sure if the volume of the drum took the vapour into consideration.

So Basically I am very confused on how to approach this and where to get my design equations for this type of drum. Most design equations i find are for Reflux Drums.

I attached an excel sheet showing a layout of my process streams into and out of my surge drum. Because of the temperature of the streams, the Acetone is in a vapour state but is a extremely small amount, so not sure how much this affects my drum volume.

I am just asking please if anyone can tell me where and what type of design equations I should use for my vessel and if I am on the right path or not ?.

Regards,
Nick

Attached Files

•  FEED_SURGE_DRUM.xls   43.5KB   182 downloads

[Art Montemayor]

Nick:

Every time a Chemical Engineer makes a decision, it should be based on need, logic, and common sense. A process design engineer only designs and incorporates those equipments that are needed for making the process work safely, correctly, and consistently. You have not furnished us with any Basic Data regarding the “surge drum”. What is it for? What purpose does it serve? What should be its function within the process? --- and the most important question: Is it really needed?

Remember I mentioned the K.I.S.S. principle in your previous thread? This is the place where you should apply it religiously. I suspect you are trying to feed a reactor with a combination of recycled Cumene and fresh, make-up Cumene. I think you are using this drum to mix, commingle, and condition the feed to the reactor. The two Cumene streams may exist at different conditions (pressure, temperature, composition, flow rate) and for that reason, the drum seems a good idea in order to “condition” the combined feed streams there and have a homogeneous feed to the reactor. This ensures a steady, controllable reaction. Am I on the right track? If so, then you should design the drum to fulfill the desired function.

What I have described above is a vessel designed to fulfill what we here in the Texas Gulf Coast conventionally call a “wide spot” in the line. In other words, all it really does is absorb the differences in the two streams and converts the combination product to a steady, smooth feed stream – a sort of shock absorber (buffer) – so to speak. Residence time is one of the primary parameters used to design such a vessel – and sometimes proprietary internals are also used by some manufacturers. However, you can’t expect to know what special internals are to be used, so you are left with only residence time to consider. And a reasonable amount of residence time should suffice. Here, I am assuming that you have SINGLE PHASE fluids – not 2-phase. If I were you, I would select a reasonable amount of time that represents an expected duration of flow variance or instrumentation total response time. I don’t know the physical size of your process or the flow rates involved, but I would guess a time increment of 5 to 10 minutes of residence time should yield a reasonable vessel size. The location of the 3 major nozzles (both inlets & one outlet), and the ability to effect internal mixing should be mechanical details of importance.

Sometimes, where you have the possibility of pressure variances, you can build in a “bladder” effect into the vessel – either by employing a membrane that separates a compressible fluid – such as nitrogen – or a natural, built-in inert gas pocket within the vessel. This literally is a shock absorber and serves to dampen any pressure “surges”.

I hope this has helped you out.

[nick_tri]

Hey all,
Firstly my apologies to Art for not providing more info on the purpose of the drum. You are very spot on though on the purpose. I am in fact using it as what you'll call it a "wide-spot". I really need to get in the habit of following this KISS rule. lol.
The purpose is to supply LIQUID cumene to an oxidation reactor. However, compared to the 9279 kg/h Cumene i have coming in, I also have a trace amount of acetone gas entering the vessel (2.552 kg/h) because of the temperature of one of the two inlet streams being above the Acetone boiling point.

I note you assumed I had single phase, but since I have a 2-phase mixture, I guess I will purge the trace acetone from the vessel. Does that make sense ? Also does this mean that I have an atmospheric vessel , because I just want the liquid cumene.

My calculation goes like this:
Using a Residence Time = 10 min ,
I got the Vessel Volume = 1.95 m3
Liquid Volumetric Flow in = 0.195 m3/min

From "McKetta" I saw that the Length(L)/Diameter(D) ratio should be 2.5-5. So I chose 3, therefore i got L=2.418m and D= 0.806m.

Now since my drum feeds a pump which in turn goes to the oxidation reactor, I figured I would provide a 1m NPSH and 0.25m vapor clearance for the trace acetone gas. This gives me a vessel of dimensions 0.806m dia x 3.668m length. Does this sound good ?


In terms of my mechanical design, how do I design the location of the 2 inlet nozzles and 1 outlet nozzle ? I plan to design for wind stresses and loads on the drum, so do i design it as a thin walled pressure vessel ? (even though it may be atmospheric ?)

I know this is alot, but i try to be as thorough as possible in all my uncertainties.

Regards,
Nick

[Art Montemayor]

Nick:

I think you are not KISSing. You seem to be overworking the problem. What you have described is a need to mix two feed streams while separating a slight amount of acetone vapor. That’s all you need to do --- plus you can assist your feed pump’s NPSHa as well as regulate the venting of the acetone by putting a slight pressure on the surge drum with a back-pressure regulator.

Refer to the attached Rev1 of your workbook and see how I have adulterated your sketch with a new version. I think a horizontal vessel works best with what you want to do. You only need simple saddle supports and no wind bracing. The internal baffles / weirs allow the two liquids to mix and disengage the acetone for venting as shown. From what you describe, I think a drum liquid residence time of 15 to 20 minutes would work well - if the drum is not too big.

For learning how to design a process drum with respect to the internal volume, go to my workbook called Vessel Volumes - which I have uploaded for free downloading on these Forums. Again, use the SEARCH feature to find it in the corresponding thread. Let me know if you have trouble.

I hope this helps you see the application is rather simple from a pragmatic and practical sort of way – hey, that’s Engineering!
 Cumene_Feed_Surge_Drum_Rev1.xls   168KB   239 downloads