Horizontal ammonia vapor drums 2

[asv80]

Can someone send me the design procedure or calculations to desing a horizontal ammonia vapor drum. I know the flow rate of ammonia, density of the vapor and liquid. But do not have any reference material with me that would describe the procedure.
Thanks
ASV

 

[Montemayor]

asv80-

Can you please be specific as to what you mean by "design"? Is it a process design or a mechanical design that you are after? It makes a big difference and only creates confusion and wastes time and effort unless you are specific in your requirements.

What is the scope of the drum in question? Is it meant to act as a separator? as a surge? as an accumulator? or is just a wide spot in the line? How can you expect help on a vessel design when you haven't stated what you are trying to effect with the vessel?

Thanks.

 

[asv80]

Montemayor,
I am supposed to come up with the size of the unit - diameter, length, riser spacing, etc. I am involved with sizing a heat exchanger that uses ammonia as working fluid. So I need the drum to act as a vapor/surge drum.
I hope this might help. If you need more clarifications please ask.
Thanks
ASV

 

[Montemayor]

asv80:

Thanks for the feedback. I think we still need additional basic data because, as you previously stated, "I know the flow rate of ammonia, density of the vapor and liquid." This implies that the vessel is handling two phases - liquid and vapor and, presumably, you want to separate them such that the liquid remains behind (drains back into the heat exchanger by gravity) and the NH3 vapor exits out overhead. Am I correct in assuming this? If so, then a 2-phase separator is called for, with sufficient volume to absorb any system flow surges. The basic method of process design for this type of vessel is the Souders-Brown relationship which takes the form:

Vmax = (k) [ (dL - dV) / dV ] ^ 0.5
where:
Vmax = maximum vapor velocity, ft/sec
dL = liquid density, lb/ft3
dV = vapor density, lb/ft3
k = 0.35 (when the drum includes a demisting section)

I've taken the liberty to copy Milton Beychok's oft repeated recommendation on the subject in many previous threads on this same subject. While the Souders-Brown relationship is favored for vertically oriented vessels, it can also apply to horizontal ones. However, a favorite horizontal vessel relationship is found in the GPSA Engineering Databook. It is based on particle size separation. Other references are:

1. Applied Process Design for Chemical and Petrochemical Plants; Ernest E. Ludwig; Gulf Publish. Co.; Vol. 3
2. Gas Conditioning and Processing; Dr. John M. Campbell; Campbell Petroleum Series; Vol. 1

There was a thread Posted on Feb 10 2004 in this Forum discussing this very subject, but unfortunately it has been removed from the archives database. Otherwise, I would refer you to it.

I hope this helps you out.

 

[mbeychok]

asv80:

Why must the drum be horizontal? Why not a vertical drum?


Milton Beychok
(Contact me at www.air-dispersion.com)

 

[asv80]

Thanks for the relationship and the references. What Montemayor assumed is exactly what is expected in the design. Do I arrive at the diameter of the drum assuming as a pipe flow in which the maximum velocity is Vmax, determine diameter for this velocity. Then determine a length from an assumed L/D of 3-5? As for references are there any design methods over the net as I would be unable to get them within the time I have.

As for the need for a horizontal construction, it is stipulated or required to be such that there are restrictions in available space.

I did look up a Gravity Separator fundamentals and design by Jekel et al on the net. It states an example of horizontal design but mentions about surge volume and ballast volume. Do you think you can throw some light on how to calculate that surge volume. If there are other design methods, could you detail it here.

Thanks for your attention and replies.

 

[mbeychok]

asv80:

I don't want to be too persistent about this point, but I don't understand how a space limitation dictates a horizontal vessel. A vertical vessel will require quite a bit less plot plan space than a horizontal vessel.

Or is it that the space limitation is one of headroom inside a building? Or what?

Milton Beychok
(Contact me at www.air-dispersion.com)

 

[asv80]

Mbeychok:
Yes I would think so. The design required is similar to a one already existing system with a horizontal drum and they want a horizontal drum in this case too. If you may elaborate the design it would help (both horizontal and vertical).
Thanls.

 

[TD2K]

Drop me a line if you like (testdog2000@yahoo.com). I have a paper on sizing horizontal separators that would help you ASV80.

 

[asv80]

TD2K,
I appreciate you for offering to send the paper. Please send it to mechengr80@yahoo.com.
Thanks in advance.

 

[NghiaPP]

If you intent to use ammoniac as cooling medium for another stream, why not integrate both heat exchanger and vapor separation in one vessel. A BKU type heat exchanger might be suitable. The shell acts as horizontal separator vessel with vapor disengaged space above the tube bundle. This space should be calculated as when you design a horizontal vessel. Use level control to control the flow of ammoniac to the shell. Vapor is taken out on the top, and maybe a blow down to avoid build up of any heavy contaminant.

 

[asv80]

NghiaPP,
I agree with you, a kettle reboiler type will be suitable for this kind of application. But, the means of adding only an extra drum to the H/X that is usually being made/stocked seemed to outweigh newer designs. However, I will keep in mind your suggestion in future cases where there is more leverage. Thanks for your comments.
ASV80

 

[asv80]

I now have a question for the surge tank design when there is only a liquid phase in the system. I have a flow rate of liquid NH3 at about 4000 lb/h and -10°F. I need to design a surge tank for this. Are there any methods to do this? Is it right to size a tank for a specified specific holding capacity of say 5 min. Please suggest the approach to this case.
Thanks in advance.
ASV80

 

[Montemayor]

asv80:

A liquid surge tank used in a process is nothing more than "a wide spot in the line". In other words, you normally have no definite, identifiable, process constraints for its need other than that you want to make sure you can maintain a steady operation of "liquid-full" downstream. If, on the contrary, you do have a defined process need for the "surge" - such as a definite quantity of incremental liquid required at certain incremental times, then you have logical basis for your surge drum sizing. I would use that basis, plus a contingency to cover suspected variances to the volume rate.

The need for a liquid surge drum is specifically peculiar to the specific process in which it is applied. All surge drums cannot, by logic, be calculated by the same general sizing formula to arrive at a logical answer. Since all processes are different, with distinct characteristics and needs, the surge drum required would be specific to that process. Surges may be identified as storage (during shut-downs or turn-arounds), process capacity variances due to production needs, variations in liquid transport equipment such as pumps, pressure differentials, etc.. Other needs may be testing, mixing, or settling of solids. It all depends on your scope of work and your process. You, as the expert in your process, are the best judge of what your needs are. I would not employ an arbitrary equation or number arrived at on an internet forum.

To try to give an example of this perspective, consider a human being's bladder: the size of the bladder depends on the obesity, the type of human, and the amount of beer he/she drinks per day. The more liquid the human body has to process within an increment of time, the bigger the bladder.

 

[asv80]

Montemayor:
Thanks for the answer. I can now appreciate the nature of the problem better. I realize that the process requirement is to store extra quanity of liquid and supply it when need arises. As such I do not have information on the fluctuations that might occur. My thought was that in cases of vapor-liquid separator drum, it is considerd to hold 5 minutes of liquid as reserve and was wondering if similar assumptions were being followed elsewhere in industrial practice. I have now to base the design on my judgement.
Thank you.
ASV80

 

[Montemayor]

asv80:

Normally, I allow a thread to fade away into the sunset after the original poster has been satisfied. In this case, I suspect there may still be a slight feeling of apprehension regarding the methodology used in designing the size of surge drums and separators using only empirical, engineering sense. This is usually a case when the engineer is limited in experience and is at the beginning phases of his/her career; I can identify easily with this feeling.

This particular example of a liquid surge drum is a case in point where the ultimate resource is good, experienced, and sound (& documented) engineering judgement. There is no substitute for sound engineering judgement and when it is documented openly and in all candor, it stands as a great help and experise for all parties concerned in an operating process. If an engineer feels he/she lacks the necessary experience or expertise, they should consult with an individual who does have the experience. The mere act of consulting with an experienced indivdual is itself a measure of the intelligence and ingenuity of the young engineer.

I would further add that in the majority of cases, such as this one, it is a wise and prudent engineer who openly seeks and obtains the help, counsel, and recommendations of his fellow peers - especially instrument engineers who know the process as well - or better than the designer. This is particularly true of vessels such as these, that affect the successful and safe instrumentation controls within a process. The case of a gas/liquid separator is one that falls immediately in this realm. The level of the separator drum is an important sizing parameter and the liquid capacity/level is often a critical measurement due to the need to either empty of retain the liquid within the vessel in sufficient time to allow operator reaction to a potential unsafe incident. I mention this because the "5 minutes" of liquid residence is not a rule of thumb nor a guideline. I would set the residence time in accordance to specific process and instrument needs and characteristics specific to that equipment's operation within the process - at all times, whether in normal or restricted operation. For a process to function correctly and safely, it must be designed with complete overlap and interchange of information between the designer and instrument engineers.

 

[asv80]

Montemayor:
Thank you very much for this piece of advice. I truly understand and agree with every word of yours. It may well have been brought out from my posts that I am not familiar with the design process. In the absence of immediate assistance by way of some experineced person who may readily offer advice at the right time, I am using this forum among other avenues as a general guidance which might lead to better understanding and possible sources of solution. I realize it is not appropriate to base designs on simple design tips, but if you may note in my posts I had asked for guidance by way of established design principles which other experienced people will know instantly when they look at a problem. I believe that is good way of learning; though I do not in any way consider worse to have to learn things from scratch on one's own. Even at this point as I have not received any real answers to my question I am still looking for it. I would appreciate anyone who will be able to point me in the right direction.
Thanks in advance.
ASV80