17 replies to this topic

[aamaterasu]

Hi,

 

I have to vaporise ethylene feed (mixture) and I am thinking of using kettle reboiler. 

 

I have to simulate it in Hysys and my tutor said it is ok to simulate it using a normal heat exchanger (Hysys does not have K shell). Vapor mixture would be sent to the reactor. 

 

However for the detailed design which I have to submit, I have to provide the calculations and right equipment to use. I am thinking of using AKU heat exchanger. Steam would be used to heat.

 

After reading the attached I have some questions.

1)Critical Pressure is needed for Mostinski's Equation. For mixture, is it ok to use weighted average from valves of single component in Hysys for calculations?

 

2)They took liquid level as 500mm and showed a freeboard of 340mm which is satisfactory (More than 250mm is ok i think). How did they arrive at 500mm? Is there any calculation and how do they ensure liquid level will stay there?

 

3)My PFD in Hysys would be recycle ethylene feed + fresh ethylene feed -> heat exchanger -> output from HX to reactor. I saw many kettle reboiler attached to a column. Is it necessary?

 

4)I previously done a design for AEU HX but in the attached, I find that there is missing design parameters like pressure drop in tube and shell / ft factor / tube side velocity / L/D ratio. They also do not iterate on the estimated U values. Are these parameters not applicable in kettle reboilers and what other design considerations I should take account of?

 

5)Also they just put a surface tension estimated of 9.7*10^-3 N/m. Is it the surface tension of the butane liquid surface? How did they get the value and how do I come up with a value for my mixture? And does the 326 *10^3 for lamda in modified tuber equation come from latent heat?

 

As you see I have a lot of questions and I thank you for taking your time to help me!

Attached Files

•  Kettle Reboilers-3.pdf   123.36KB   252 downloads

Edited by aamaterasu, 01 March 2016 - 10:24 AM.

[Pilesar]

My comments to your kettle questions:

1) Since you have Hysys available, why not use the pseudo-critical pressure it calculates for the mixture? If you have to calculate a critical pressure yourself, Lee and Kesler correlated critical pressure of hydrocarbons as a function of boiling point temperature and specific gravity. (reference 1976 Hydrocarbon Processing article "Improve prediction of enthalpy of fractions")

2) Liquid level in a kettle is often maintained by an internal weir to make sure the bundle is always submerged. This is usually a minimum of 2 inches with additional submergence given to allow for foaming. The free board is the disengaging space available where the vapor can separate from the liquid droplets produced by boiling. The liquid entrainment that can be tolerated depends on where the vapor is discharged after leaving the kettle.

3) Kettle boiler does not have to be attached to a distillation column. For vaporization service, you would normally not call it a kettle reboiler, but a kettle vaporizer.

4) The kettle is a heat exchanger, so U value, and other parameters are important. The disengaging space required depends on the amount of vapor generated. The total volume of disengaging space required can then be divided by the length of the bundle to find the required cross-sectional area above the liquid level. You should have a minimum height of at least 12 inches above the liquid level.

5) Surface tension is a liquid property. You can use the value from Hysys or research this yourself. If you find this for the pure components, you can sum them proportional to the mole fraction of the components in the liquid phase. I do not recognize 'modified tuber equation', but lamda is generally used for latent heat.

6 (bonus) It is generally better to have a kettle boiler too large than too small, so give it plenty of disengaging space and extra length if you can. There are many additional details that are important if you are specifying the exchanger for purchase, but these details are probably not needed at your stage of the design.

[Art Montemayor]

My personal experience with kettle reboilers and kettle vaporizers agrees with essentially all that Pilesar points out - with only one exception, which I think should be considered:

 

A kettle heat transfer operation is essentially a basic submerged vaporization application.  This type of operation is a unique and special application in heat transfer and is one where the application of the Overall Heat Transfer Coefficient ("U") is usually an academic one and not practical in actual practice.  The reason behind the possible unimportance of the "U" factor in this application comes about due to the submerged vaporization phenomena known as "nucleate" boiling - the formation of surface bubbles on the heat transfer surface.  The opposite of this effect is the "film" boiling phenomena - which is to be avoided if the heat transfer is to take place efficiently and consistantly.  Instead of a "U" being an important factor, I have always relied on applying a conservative heat flux (Btu/ft²) value to the applicable tube bundle employed.  I have used heat flux values of 6,000 to 10,000 Btu/ft² as design figures.  Kern, in "Process Heat Transfer", treats this subject in detail in his Chapter 14 - Evaporation and makes note of the negative insulating effect that takes place during film boiling in relatively high heat flux applications.

 

A. Kayode Coker, in "Ludwig's Applied Process Design for Chemical and Petrochemical Plants" also details out the importance of this phenomena.  I believe this is an important design criteria to take into account during the design of any submerged boiling (or evaporation) application.

[Pilesar]

I designed a kettle exchanger to vaporize 500 psi ethane with 30 psi steam using HTRI software. (This eliminated the necessity for me to use equations to manually calculate heat flux values or heat transfer coefficients in the exchanger design.) For my ethane vaporization service, there is a high temperature difference between the tube metal and the boiling liquid which means high heat flux occurs. The heat flux in my kettle calculated to about 30,000 BTU/ft2 which was about 30% of the calculated critical flux. It is my understanding that the ratio is what is important instead of the value of the heat flux itself and this response is based on that understanding. I will check Kern and Coker when I get access to them later. My vaporizer bundle is only eight rows high (60 tubes) and every check I made pointed me away from concerns about vapor blanketing the tubes. (The kettle has not yet been commissioned, so I do not have proof my design works!) As bundle size increases, the bubbles in the upper half of the bundle have a more significant impact on heat transfer. Heat flux is still directly proportional to temperature difference, so delta T is still an important parameter. But there is no effective heat transfer in the film boiling region so heat flux is usually used in charts to predict the heat transfer coefficients for boiling services. It still should be referenced as a check to make sure that you stay well away from the unstable region and keep the heat flux less than 50% of the maximum flux. I admit I end up trusting the software to choose the correct equations I need as I do not have as much experience in exchanger design as some. Thank you Art for raising an important point. It pushed me to reviewing my design yet again to make sure my design is sound. I bootstrapped myself into heat exchanger design and I really enjoy it, but there are depths I have not fathomed. Boiling heat transfer equations make my brain hurt.

[aamaterasu]

Thank you very much for the very insightful advices. However after some discussion and reading, I heard that kettle reboiler is seldomly used in practice and it was rumoured amongst the students that we should do the design on a vertical thermal syphon instead (I do not know why). 

 

A visit to an industrial plant (Producing VAM which is the product of interests for our design project) also showed that they are indeed using a vertical thermal syphon instead of a kettle reboiler. Please advise on whether I should continue with the design of the kettle.

 

*On page 247, the book mentioned that thermosyphon reboiler also cannot be used as a vaporiser so I am confused now. https://books.google...nepage&q&f=true

Edited by aamaterasu, 04 March 2016 - 04:59 AM.

[Pilesar]

I suggest choosing the technology that you can execute the best. Kettles have successfully been used to vaporize ethane in industry so they are not 'wrong' even if a thermosiphon might be 'better'. If I were your professor, I would not disapprove of either selection. To my mind, a kettle would be easier for you to design as thermosiphons can have tricky hydraulics that must be considered. But school is for learning so pick a technology and learn it well. If you have a 'go-by' example to follow, then that should make the decision easier. One thing to consider is that your ethane feed is probably not 100% ethane so you should take account of that in your design. Composition variation may be one of the reasons the industrial plant you visited chose thermosiphon technology. If your project calculations are accurate and your presentation is clear and reasonable, then I doubt your selection of vaporizer style will matter to your project grade.

[aamaterasu]

Hi thank you for your advice. I am currently in the mist of reading up more design considerations for the kettle.

 

When I am simulating the kettle in Aspen EDR, I received a warning message saying something along the line that the liquid would not be able to fully submerge the tube bundle. Is there any way I can overcome this problem?

 

Also is tube side velocity/pressure drop calculations important too?

 

Thank you!

[Pilesar]

EDR will size the exchanger for you. I recommend that path as a starting point. If your kettle has an internal overflow weir, the weir height and tube layout are set so that the tube bundle will be fully submerged when in operation. (The liquid level will be higher than the weir.)

 

You are condensing steam on the tube side so the velocity will be pretty fast. It will be difficult to keep the inlet velocity low, but try to get it down toward 100 ft/sec. You may have to settle for higher velocity, but 200 ft/sec is extreme and you should be able to do better. Pressure drop on the steam side is probably not an issue if you get the velocity down. The condensing pressure may be so low that you need to pump your condensate away.

[aamaterasu]

Okie thank you very much for your help! By the way, I am trying to heat a 30deg acetic acid feed to around 200+ deg to vaporise it. Is it operationally feasible for it to be done using just kettle reboiler?

 

I read somewhere kettle is only used for latent heat transfer. When I think about it it also make some sense. Do you actually use the heat to boil the feed or do you use it to raise the temperature? Its kind of confusing

 

This is what my tutor told me:

"Just for example, suppose I want acetic acid vapor at 1 atm. I can give the heater specifications as 1 atm (BP-118 deg C), and vapor fraction as 1. It will give me the Q value. Now suppose I want superheated acetic acid (1 atm and 130 deg C). I can give this as specification in heater and get back a Q' value. But is this physically possible in a kettle reboiler. No. Since the vapor and liquid are always in equilibrium. So you will get maximum 118 deg C (considering the tubes is completely submerged in liquid). If you have to take it 130 deg C, then you will have to add an additional exchanger which will heat the vapor from 118 deg C to 130 deg C. And the combined heat duty of both the exchangers is given by the Q' value."

 

Does this mean I have to preheat ethylene using a normal heat exchanger to 200deg(aq) and then send it into a kettle reboiler where it vaporise?

 

Also my liquid feed coming into the reboiler is around 8bar. Should I reduce the pressure using a control valve so that liquid and vapour separation would be better?

 

Thank you so much!!!

Edited by aamaterasu, 11 March 2016 - 11:47 AM.

[Pilesar]

You do not superheat vapor in a kettle because the bundle is under the liquid surface and not in contact with the vapor. Physically, you can feed subcooled liquid to a kettle and boil it, but this is usually only done when the liquid is pretty close to the boiling temperature. When the feed has many degrees of subcooling as you describe, additional exchangers are probably warranted. The design of the heating train is often based on economics considering the cost of energy and the cost of capital. Efficient energy usage means that you use hot process streams you need to cool to add temperature to cold process streams. Only after you exhaust the practical uses of transferring heat between the process streams do you rely on using utility streams to finish the cooling and heating. Another consideration is that heating a single phase stream is better done with another type of heat exchanger. The total heat exchanger surface area may be less when the exchangers are optimally designed for single phase like heating a liquid or superheating a vapor. It is simplest to design separate exchangers for liquid heating, boiling, then vapor heating because the results are more predictable. 

Reducing pressure before vaporization is probably not a good idea, but may be required depending on the temperature of your heating medium. You will need pressure to move the vapor to its destination. Evaluate your flowsheet in advance and note that there are some places where you can fix the process conditions. If you can design your process to get to those places with minimal additional power or utilities, you will likely have an efficient plant. Typically you want to get as much use as you can from the feed pressure as it can really make the difference in whether the process is profitable or not. You will have to lose some pressure across control valves, but pressure is not something to waste as long as you have use for it.

[aamaterasu]

Hello after doing some preliminary calculations,

High pressure steam 250 to 220 water

Feed 80C liquid to 205C Vapour

 

My kettle has a 

shell diameter: 1.382m

Bundle Diameter (u-tube): 0.691m

Heat transfer area : 110.8m2

Tubes: 152 U-tubes

Heat Flux: 87.8 KW/m2

Max heat Flux by zuber: 190KW/m2

U: 1021w/m2c

Steam inlet velocity: 15.6m/s

 

Are these numbers making sense?

What are some other calculations I have to do? I tried to calculate pressure drop across tube side but it requires reynolds number calculation regarding two phase flow. I read it up and looks really complicated so I just gave up. Is it necessary?

 

I also have to do heat integration for the entire plant. Is there are easy way to do it? I am only in charge of the heat exchangers and my other group members are doing distillation/absorber column. Some of my friends are doing pinch analysis using Matlab and if theres no easy way to do it I guess Ill just attempt to connect any heat output stream to heat input stream.

 

My design project is finally coming to an end soon (love-hate relationship with engineering)...thank you guys so much for your help!!!!

Edited by aamaterasu, 18 March 2016 - 11:53 AM.

[Pilesar]

Nothing jumped out at me as unreasonable in your kettle. For final design, you will need kettle length, a control system for the process and steam, weir height, tube layout, materials of construction, design pressure and temperature, nozzle sizes, etc. For a student project, I would be satisfied with exchanger area and duty in an equipment list, but you may be able to impress your professor with more specifications.

 

Heat integration should be done before exchanger design. If you integrate heat after exchanger design, you will change the design basis for the exchangers.

 

Pinch analysis can be done using special software, but you can do much on paper. I mentioned the principles of heat integration in my earlier post (see the section beginning 'efficient energy usage means...') Pinch analysis helps determine which process streams can be used for heating or cooling and the optimal configuration for the energy integration. Pinch is really pretty easy... list all your process streams that you are heating and all the process streams that you are cooling. List the heat duties and the temperatures. Match the streams so that you get the most energy efficiency. There are constraints to consider such as pressure drop and exchanger approach temperature. It is probably worth while reading a couple of written references on pinch technology to get a feel for it. It may look like a lot of work, but it really won't take you but a few hours to master the concepts well enough to apply them in a crude way to your project. I would not bother with Matlab. If I were a professor, I would be more impressed with hand work well explained than with inscrutable software results.

[breizh]

Hi ,

You may find some interest reading these documents , you already got a lot from our colleagues.

 

Good luck.

 

Breizh

Edited by breizh, 20 March 2016 - 06:45 AM.

[aamaterasu]

Hi thank you for the very informative article and all the advices that were given! I read up on pinch analysis and did some pencil and paper work.

 

I used an excel file available and have currently generated the composite curve and the "grid diagram". However I have one question, do I have to include all the duty in (inclusive of duty from compressor/pump) ? I included all duty into the file (duty from heat exchanger, pump, compressor,decanter) except reboiler and condenser. Or only duty from process stream are required (heat exchanger)?

 

This is because it is very easy to visualise energy transfer from process stream using an heat exchanger. However for energy needed/generated from other sources, it is an energy stream and not an process stream. I was thinking it might be possible to transfer these heat energy into a process stream (i.e. converting water to steam to store the energy) but am not sure on how we can "collect" the heat energy. I hope I am clear enough with the problem/issue that I do not understand. 

 

With the grid diagram generated. I have matched all the streams below the pinch and above the pinch with heat exchangers such that no duty is being wasted . A quick paper calculation has shown that the minimum heating and cooling utilities are indeed very close generated by that of the excel file. Is this the correct approach?

 

Thank you very much!!!!!!

Edited by aamaterasu, 22 March 2016 - 08:03 PM.

[Pilesar]

It sounds like you are on your way to Pinch Analysis of your system. Aren't you glad you did not spend time programming Matlab for this? You can include only the heat exchangers in your Pinch Analysis for heat integration. A full exergy study may be valuable, but is more theoretical and probably beyond your charter to design a plant. Take your time in the early stages of the design to eyeball your system to see if rearranging the process flowsheet would be beneficial. If you have a go-by flow diagram, someone has likely already done some overall optimization so don't spend much time on that. Do try to conserve compressor power by increasing pressure of liquids instead of gases where possible. Compressors greatly affect plant cost. Don't rush through the early stages of your project as this is the place where decisions can make a big difference on outcome. It is very inefficient to change project scope after equipment design has started. Engineering work is not just pushing a pencil and calculating, but consists also of thinking and contemplating.

 

For your pinch study, remember that there will be a 'gap' between stream temperatures to account for usable temperature approach between streams.  Suggestions for you to consider: I suggest putting your Pinch Analysis of the ideal system into an addendum of your report. There may be differences between the ideal and the practical which will bite you when you design your exchangers. For your purposes, I suggest you resist the temptation to maximize energy savings, but leave a nice comfortable operating range (maybe using a minimum pinch of 30 F approach or greater) for your process-to-process heat exchanger design. If you over-optimize, you run the risk of having to re-do all your heat integration later if you need to revise an exchanger duty. Ignore small duties and concentrate on the larger ones where you get more energy savings. The goal is for the plant to be able to operate and not to squeeze every last BTU out of the heat train. If your imputed cost of energy is low (as currently in the real world) then capital cost may be much more significant than energy cost. If you include your pinch analysis in your report, you show your understanding of the concepts and the possible energy savings of continued optimization. In the project design, waste some duty to simplify your work and provide a more flexible operating plant as it may save you some frustration later.

[Omar U]

Hi there,
I am designing a kettle boiler for the purpose of cooling a styrene reactor effluent stream from 610C to 240C by the vaporization of sat water at pressure 15 bar.

1- I'm searching for a go-by to follow but haven't found any but Coulson's example and i'm wondering if it is enough or not for doing a detailed design and whether there are any avalible solved examples to follow?!

2- For the tube side heat transfer coeffiecient and pressure drop calculations should i follow the same procedure as that done in normal shell and tube exchangers or something will change?

3-What is the effect of the presence of u-tube bundle in my design instead of the normal tubes?


If there is any information that you need in order to give a more helpful answer please mention it

Thanks in advance

[Art Montemayor]

Omar:

 

You are a student, I presume, and you fail to state what KIND of design you are undertaking: process or mechanical?

 

The best helpful advice I can give you at this juncture is: READ THE ABOVE THREAD and you will become aware of the obvious: your topic is totally different from that of this thread.  You show that you are not familiar with the difference between a REBOILER and a HEAT EXCHANGER.

 

Start your own thread instead of injecting your specific need to another member's thread.  This is analogous to hijacking another member's thread.  I will delete this posting as soon as you start your own thread.  This should happen quickly, so other members don't jump on your posting, forgetting that your topic is not in concurrence with this thread.

[Omar U]

Thanks Art Montemayor for the advice,
I have read the thread and thought that my post is relevant as the design procedure should be similar with few differences.

I apologize for this and I'm waiting for your help in my own thread.