10 replies to this topic

[Teknas]

Hi,

I am sizing a steam knockout drum located at the steam compressor inlet.

Knockout drum sizing requires the liquid fraction in the total vapor flow. But in this case, the steam comes directly from the steam header, i.e. I have no information about the liquid fraction in the flow. Suction pressure to the compressor equals the header pressure.

How do I size the drum ? I have been advised to assume a liquid fraction of 3 - 5 % in the steam flow. Is this assumption good enough, and what would be the basis ?

If there is any other of going about this sizing please tell me.

[Qalander (Chem)]

Dear

why don't you use 'search' option of the forum?

As this topic is a greatly discussed one, Lot of useful info you should be able to obtain.
Wish you best of luck.

[ankur2061]

Sanket,

I can't think of sizing a KO drum without having the liquid fraction data available as an input to the sizing calculations.

As far as assumption is concerned, at some stage or the other somebody might ask you to validate your assumption and that is where most design engineers find themselves in a soup. I have gone through this type of situation at the begining of my career many times and I have always found it to be a very dicey situation where I assumed ('ass u me') something and later could not validate the assumption.

In my opinion, if a superior or client asks you to assume a particular data as an input for any design which might be difficult to validate then the best way is to have it on record (design basis document, minutes of meeting, e-mail confirmation). That way you are protecting not only yourself but your employer (engineering consultant) and everything is in black and white, so that if there are any accidents/mishaps which may lead to legal proceedings, the legal authorities know the origin of the data and responsibility can be pinned down.

With experience over the years and looking at the system a veteran engineer can also make some kind of a judgement about the veracity of the data. Sometimes just looking at the numbers, an experienced engineer can make out if the data is outright crap, and I can personally vouch for it.

In a nutshell try to get some real data by insisting on its importance. If you can't, have things on record so that finger pointing and heart burn can be avoided at a later stage.

Regards,
Ankur.

[Teknas]

Hi,

Thank you Ankur for your suggestion. What you say is absolutely correct, but then there has to be some criteria for design.

Even when a compressor vendor will design the knockout drum for the 1st stage, some assumption or thumb rule has to be followed. I would like to know the basis for that. As you rightly mentioned, the 3-5% criteria seems to have come out of thin air and no one is sure from where these figures came. The practical problem arising out of this is that the k/o drum may be oversized, or even under sized.

I tried to work on a concept, strictly for steam, that usually steam traps are placed at a distance of 30 m. So assuming the piping length of 30m, find the pressure drop. Then based upon the difference in enthalpy and multiplying by the steam flowrate get the total heat loss and then divide by the latent heat to get the amount of steam condensed.

But this gives a very small condensation.

I would be really grateful if some specific guide line is available for such a scenario.

[Qalander (Chem)]

Dear

Firstly I really apologize for a very general response due to my haste..

Secondly, is this correct for me, to understand that you intend to compress steam

to obtain what? some high pressure steam or condensate etc. I feel lost here.

Sorry if I am mistaken can you urgently clarify for me to understand the issue correctly please!

[Srikrishna Chaitanya]

Hi , here are some ways to clarify your doubts

1. I f liquid exists in your steam , you can find it by calculating vapor fraction , you know Pressure & Temperature from any steam table or simulation software you can find vapor fraction .
2. Let us assume there is no liquid in your steam, you would be sizing KOD by using some guide line like SHELL DEP /API. In these guidelines you may get minimum values for Gas load factor etc, try to refer whether some thing is there on minimum liquid flow rate.
3. While doing any equipments design , customer will definitely give Process design criteria / design basis , please try to look in to that , you may get control times , max / minimum liquid content etc.
4. I feel 3-5 % of liquid (vol /vol )assumption is quite high, clarify with experts


If you can tell why you are doing this design , as a part of learning / normal job and how important is your design(For fabrication or estimation etc.) then people can guide you better



Regards,

satish

[ankur2061]

Sanket,

The arbitrary assumption of 3-5% liquid fraction seems to be way off from reality. As I had mentioned in my earlier post you need this as an input. From a practical view point, the amount of condensation in a well designed (pipe routing) and well insulated steam header is negligible. For heat loss calculations from insualted pipes I have put a previous posting whose link is as below:

http://www.cheresour...?showtopic=6248

Another post relates to jacketed pipe where the amount of steam required to be fed to a jacketed liquid sulfur line due to heat loss from the pipe is as below:

http://www.cheresour...mp;hl=heat loss

See if you can actually calcualte the heat loss from the pipe and determine the amount of condensation. A simulator would really help, if you can lay your hands on one (HYSYS). However, don't assume anything without any sound reasoning and thinking.

Regards,
Ankur.

[djack77494]

I have never heard of a steam compressor so that piqued my interest. Anyway, to design a K.O. Drum, you do not need to assume that any liquid phase is present. By sizing, I assume you mean to determine a reasonable diameter and height for the drum. Just using the Brown-Souter Equation, you can calculate the vessel's diameter. There are various, often proprietary guidelines for setting the vessel's height based on reasonable allowances for vessel internals as well as liquid surge. No liquid phase merely means no need for liquid surge volume. Carry on.

[Teknas]

Hi,

Thank you for all your replies.

I have decided to let go the criteria of 3-5% liquid fraction which anyway had no basis. The backup for this is the quality of steam, which is 1. The knockout drum will be designed based on the available steam quality which means no liquid phase needs to be considered. This seems to be the most logical and sound starting point for the design. Since the client has not specified anything, it is important for me to fix the design basis based on the available data, and get it approved from my higher ups as well as the client.

I have a spreadsheet for sizing the knockout drum based on the Souder-Brown equation, so I will be using the same. Additionally I also referred to the GPSA guidelines for the knockout drum design. So, from my side my groundwork is almost done and seems to be sufficient.

Now to just clarify a few things which some of you has asked:

This is an ongoing project.

We are recovering heat from a column overhead, by by-passing an existing air cooler, to produce some LP steam. This steam will be compressed through a 3-stage centrifugal compressor and exported as MP Steam to a neighboring plant.

Thanks once again for your prompt answers.

[Teknas]

Dear Mr. Qalander,

Even if your advice was in haste, I did take the effort to search this forum, but could not find the answer to my query. Thank you for your effort.

[Art Montemayor]

This comment is not related to Knock Out Drums, but it has to do with Steam Recompression - a subject I have been interested in the past and one I try to follow with a curiousity that won't abandon me.

I have recompressed steam with a jet venturi nozzle. However, I have always suspected that an "old-fashioned" reciprocating device would also do the deed. I have noted that the Department of the Energy (USA) has written a lot of advertising on the subject and even does a quick energy savings calculation. This is very interesting to me, as I have in the past often been forced to vent a lot of 5 to 25 psig saturated steam. I hate wasting good energy and especially having to consume and treat additional water.

A compilation of information on the subject is attached for anyone's interest. I believe the German manufacturer identified in the Workbook does the recompression with a reciprocating compressor.

I will not comment on a Knock Out Drum since I have already done this, ad nauseum, in past threads which can be obtained by using the SEARCH feature.
 Steam_Re_Compression.xls   215KB   185 downloads