6 replies to this topic

[cjohn]

Hello Art and All,

I am working on the design of a partial condenser to separate a vapor stream of water and an organic that boils around 200°C. The cooling medium is pressurized and pumped condensate. The water/organic vapor stream has the opportunity for getting a slug of high viscosity material, so we want to put this on the tube side for easy clean-out with the cooling on the shell. Aspen modeling shows that a vertical shell and tube knockback/reflux style will be about 4 times more expensive than a horizontal shell and tube for the same duty. This is mainly due to the flooding velocity of the inlet vapor and the subsequent need for 10 times more tubes. Some fellow engineers have said that a horizontal tube-side design will not work well because the liquid will fill part of the tube and reduce surface area for condensation. The second concern is that the vapor exit velocity from the tubes is greater than 25 ft/s.

I believe that we can design it to the appropriate size as a horizontal and still see significant cost savings over a vertical, but I do not know just how much bigger to make it. One idea was to angle it slightly off the horizontal and make sure that the tubes were sized for self-venting flow. If not, we could increase tube count. Same opportunity for the vapor velocity, we could increase tube count to reduce velocity.

1. Do we need less than 25 ft/s vapor in a horizontal tube so we don’t entrain, either in the tube or at the exit?
2. What else do I need to take into account?
3. Is there anything inherently wrong with a horizontal partial condenser?

I appreciate any insight/experience that you can share.

Note: Sketch attached. We will be controlling the outlet vapor temperature using a valve on the condensate/cooling line. The vapor stream is the overhead of a reactor. The reactor is at 150 mmHg vacuum; the vacuum pumps (LRVP) are downstream of the partial condenser, total condenser, and entrainment separator. The composition of the overhead stream will vary with time from around 50% organic to 90% organic, so we have a varying load as well.

Thank you,

CJohn

Attached Files

•  Partial_Condenser_1.xls   23.5KB   354 downloads

[Art Montemayor]

cjohn:

Thanks to your Scope of Work explanation and your schematic, I'm able to understand what you are designing and it's inherent trouble spots.

Attached is my Rev1 of your workbook with my comments and notes.

Basically, I've done what you are up against. I've worked in Kettles operations, making custom products in batch, kettle operations. I've done both the horizontal and vertical condenser design. I am more partial to the vertical design in this type of operation because my experience shows that it is more economical (from a grass-roots point of view). However, in a retro-fit, it might cause you some headaches when it comes to head room.

In either configuration, you must ensure that you have a maximum design that conservatively can handle the flow rate variations that are occurring throughout any batch. Therefore, in your application, you are going to have an excess of tubes at the outset. That means that temperature control of the outlet vapor is an optimistic view. What I've found to be the general case is that by selecting the coolant temperature and ensuring abundant coolant flow (analogous to a "bath") you can fix the temperature approach of the vapors. That way, the controls are much easier and more practical. As fouling increases in the tubes, this may require coolant temperature changes, but operators can adjust this as needed per their experienced operation.

One of the key advantages of a vertical condenser (if you have a fouling vapor in the tube side) is that you can maintain it much more efficiently and with LESS CLEAN-OUT COSTS than a horizontal unit. Look at the sketch I've supplied in the workbook. If you have the head room and configure it right, you may just need one platform at the top "A" channel head to rod out the tubes. The sludge and solids removal is much easier and quicker with a vertical unit. Maintenance is much more efficient, quicker, and cheaper. The "wiped-film" type of condensing that takes place in a vertical unit should be more efficient than a horizontal unit that must suffer the liquid inventory at the bottom section ("belly") of each tube. This liquid tube inventory effectively nullifies a percentage of tube area as being available for condensing service - and consequently increases the size of the condenser.

You definitely have to slope a horizontal, tube-side condensing condenser. How much depends on experience with the operation and fluids involved. With sludges and solids involved, this slope could quickly approach a total vertical orientation. Maintenance and experienced engineering should be consulted on this type of application.

I would not allow nor design for the size of pressure drop you are noting across the partial condenser (60 mm Hg). I consider this too high and self-defeating in a partial vacuum reactor application. I believe you can design and obtain a partial condenser with less pressure drop - possibly half of that.

I don't have any faith in a simulation program's ability to design a condenser - especially a partial condenser - with any degree of accuracy. I prefer to leave that to proven and reliable programs like HTRI instead. Aspen may be fine as a process simulator but to expect it to be the simulator for all seasons is, in my opinion, a bit too much. That's why I challenge the 4x size of vertical condenser being predicted by Aspen. Agreeably, I don't have the algorithm, input, and simulation results in front of me but my experience tells me that it couldn't be that much of a difference. In fact, I would expect the reverse - depending on the way the simulation is laid out and inputted.

Basically, I would expect the vertical partial condenser to be much more efficient and reliable. However, the trade-off would be the head room required and/or the costs to achieve it. There is always a price to pay; no free rides or free lunches.

I hope this experience is of some helpful input.

 Horizontal_Partial_Condenser_Rev1.xls   240KB   395 downloads

[cjohn]

Art,

Thank you for your thorough response and consideration.

I was confused on why you believed that the vertical condenser could be made for the same price as a horizontal, until I saw your picture. The key point was your entry of vapor from the top with a down-flow of both vapor and liquid, a prospect I had not considered. I think that this could work well for us and the benefits for cleaning and reduced fouling would be great. The vertical that Aspen had produced was based on vapor flowing up through the tubes with the liquid flowing down the tube walls. With this type of vertical (knockback/reflux) style condenser, one of the major considerations is vapor flooding velocity. We know this from experience with similar installations. The 4x cost on the vertical was due to needing much more cross-sectional area than with the horizontal, since you do not have to minimize your vapor flow rate quite so much. Therefore, I believe that the horizontal design that I was looking at would be essentially the same as the down-flow vertical that you suggest.

Thank you also for correcting the flow direction on the coolant side. That definitely needs to allow air to escape during a normal startup. Additionally, I will get the pressure drop under control.

We are currently working on characterizing the maximum load conditions, which you have obviously had to deal with in the past. I had not looked into the potential problems of the outlet temperature control over the varying flow rates and you bring up an important point. For our process we want to pull out as much water as possible through the overheads. This means that, with the highly varying duty, we may need to lose additional organic with the water during certain parts of the reaction.

I would like to clarify your point on temperature control. You recommend maintaining a constant temperature approach as a good method to keep the water in the vapor. Do I understand you to mean that we would control the temperature difference between the vapor inlet and the cooling outlet by varying the cooling flow rate? I see this as a “feed-forward” method that would help to handle the changing duty. This seems better than controlling the outlet temperature with the cooling flow rate, which would be more reactive as a control scheme and may not keep up with the changing duty.
--> Please let me know if this was your intent or if you were going a different direction.

I am playing around with the Hill equations. They appear to be much more conservative than those that Henry Kister recommends (Moore and Sewell, separately) in Distillation Operation. I will have to read more of Hill's article to understand better where the differences are.

I appreciate your experience and your willingness to share.

It also looks like you have two excellent reasons to wake up each morning. Enjoy grandparenthood to the max!

Thank you again,

CJohn

[Art Montemayor]

Cjohn:

Because of your capable and professional-level of query that you posted - complete with an accurate schematic and explained details - we were able to communicate accurately and efficiently in a matter of 2 simple posts.

This may not seem as much of an achievement to a polished professional as yourself, but it stands out in our Forums and I make note of this success for the benefit of our members who may be reading and following this thread. I hope they capture the reward and success of good, accurate engineering communications from the very outset and the positive results that come out of that.

You are correct in interpreting my response and recommendation for controlling the temperature of the outlet vapor on the partial condenser. I think you and I see eye-to-eye in that respect. You certainly want to eliminate as much heavy organic going up with the steam vapor to the final condenser. I suspect that the heavy, refluxed organic is the "cash crop" and if this gets into the final condenser because the outlet temperature out of the partial condenser is allowed to migrate too high (the reactor vapor product rate increases), the result will be lost product or diluted product that is very costly to recover. In the perfect design case, only water vapor and/or lights are going up to the final condenser that is really there as an environmental control. Am I correct? I also suspect that you can control both the temperature and flow rate of the coolant used in the partial condenser. This coolant, I would presume, is really pumped and cooled in a closed, circulation cycle and dedicated to service primarily the partial condenser.

Thank you kindly for your compliments on my beautiful granddaughters. You are absolutely correct. They are the main reason for my waking up everyday with the expectation of seeing their beautiful faces smiling and loving me as much as I love and cherish them. Life doesn't get any better than that and I have the Great Lord to thank for the gifts I have received.

[cjohn]

Art,

You are very correct about accurate and concise engineering conversations. From my point of view, anything I can do to make your "job" easier will make my job easier, whether you are a colleague, a vendor, or a very helpful retired grandfather.

Thank you for your response to my clarification on the control scheme. This will be relatively easy to implement. I have not given you the full story due to intellectual property concerns. Suffice it to say that a little organic over the top is not as bad as water going back down. If we can provide a much greater level of control with this scheme than with others, then our downstream environmental system will have much less work to do.

I plan to move forward with this approach to the project (vertical downflow and constant approach temperature control) and also to test the control scheme on a similar installation. As I am able to get results, I will post them here to let others know how it worked out.

Take care and enjoy life!

CJohn

[Art Montemayor]

Cjohn:

I know all too well the concerns when one is dealing with proprietary information. Nevertheless, with your professional handling of the information, we were able to receive an accurate "picture" of the possible problem(s) and a quick, effective communication took care of everything else. All that is left now is some nice, detailed engineering on the condenser and the internal baffle(s) in the bottom bonnet that allow for efficient vapor-liquid separation and also permit removal of fouling matter that will collect in that section during maintenance. If the details are done well, you should wind up with a unit that not only allows for quick, efficient, and low-cost cleaning but it should hardly ever require dissasembly - except for some safety or corrosion inspections in the far future.

The best of luck to you and the Bears/Cubbies.

[jcazenave]

Cjohn:

Sorry but when you mentioned Aspen which software from Aspentech do you mean? Aspen+ or the Exchanger software like Tasc, Hetran or Shell&Tube?

thanks