15 replies to this topic

[charlie_KC]

Hi all,

 

Hoping that someone may be able to push me in the right direction. I'm attempting to perform design calculations for a simple and small shell & tube heat exchanger, for an academic project (simple in the sense of excluding calculations relating to nozzles, front/rear heads etc). However I have faced an issue regarding exceedingly large pressure drop across my baffle windows.

 

Pressure drop accounting for cross flow and end zone is approximately 10kPa (reasonable), however the pressure drop across the baffle windows is calculated as 249kPa- much too high. I can't understand why it's such a large drop through the windows, and how I can reduce this?

 

I increased the baffle spacing to reduce the number of baffles and increase the crossflow area, and I also increased the baffle cut to reduce the velocity through the windows. These geometrical variations have allowed me to reduce the window pressure drop to 144kPa, which is a good reduction, but still too high. 

 

Can anyone provide advice on how to reduce this further without impacting the heat transfer coefficient too greatly? I have no experience in heat exchanger design, and most of what I have learnt is from Heat Exchanger Design Handbook (VDI-Verlag GmbH) and Coulson & Richardson's Chemical Engineering Design, so I am unsure if I am missing some fundamental principles when it comes to geometry ratios? Otherwise, the issue could potentially lie with my fluid input parameters (which can also be altered but I want to avoid phase-change).  

 

The current design has a 20% heat transfer area over-design, so I do have some wiggle-room in terms of impacting the heat transfer coefficient. I have also attached my design calculator (sheet titled "Numerical"), if you wish to view the current input parameters (rows 6-16). Apologies for its state, it was never my intention to share the calculator.   Calcs #3 - S- 290, T- 20.xlsx   42.11KB   10 downloads

 

Any contributions would be well and truly appreciated.

 

Sincerely - 

a mechanical engineering student who is in way over their head, but recognizes it's too late to turn back now

[breizh]

Hi,

You may want to read this paper from one of our member .

Edit : a new resource added .

Good luck

Breizh 

[Pilesar]

I think from your spreadsheet that this is to be steam condensing on the shell side. Your pressure drop is high because your steam is not condensing. With steam, the heat transfer takes place at essentially a constant temperature and constant pressure. Find the pressure of the steam in the saturated steam tables and you will find the temperature where the heat is transferred. (The LMTD will use that temperature and not the superheated steam and subcooled condensate temperatures.) Estimate the total steam flow from the exchanger duty and the heat of vaporization from the steam tables. You do not force the steam flow to be some arbitrary rate... you calculate the steam flow based on the duty required. Your liquid on the tube side goes from 20 C to 165 C. Is there supposed to be a phase change? If so, you must account for that also!

  Your exchanger design is a mess. I suggest you start over. The more you try to fix the existing work (and you obviously put a lot of work into this), the behinder you will get. Try to find a worked example of an exchanger with steam condensing on the shell to use as a go-by. I am sympathetic as I made huge mistakes in academic projects. The goal is to learn if you can while also trying to make a good grade. After you calculate the flow rates and heat duty, then there are more tips you likely need for equipment design. Much of your calculations seem unnecessary and are probably adding to the confusion. 

[Bobby Strain]

I never designed a heat exchanger in school. Or any other equipment that I can recall. We learned the fundamentals. Designing stuff takes away valuable time from learning. Maybe the curriculum is designed to satisfy what industry thinks they want. I think higher education should get back to teaching the fundamentals. And eliminate all the process simulators from the curriculum! Art will probably agree. We should also eliminate the use of spreadsheets. Nobody learns much by using them. And it encourages bad habits which stick forever. Who can remember what they did with a spreadsheet after 6 months? We were introduced to FORTRAN briefly, but didn't dwell on it. Curriculum should include at least 2 courses of computer science. And a bit of history with the worlds dictators and despots. They will always be around if we allow them. Ben Franklin had it right.

 

Bobby

Edited by Bobby Strain, 21 March 2022 - 10:44 PM.

[charlie_KC]

I think from your spreadsheet that this is to be steam condensing on the shell side. Your pressure drop is high because your steam is not condensing. With steam, the heat transfer takes place at essentially a constant temperature and constant pressure. Find the pressure of the steam in the saturated steam tables and you will find the temperature where the heat is transferred. (The LMTD will use that temperature and not the superheated steam and subcooled condensate temperatures.) Estimate the total steam flow from the exchanger duty and the heat of vaporization from the steam tables. You do not force the steam flow to be some arbitrary rate... you calculate the steam flow based on the duty required. Your liquid on the tube side goes from 20 C to 165 C. Is there supposed to be a phase change? If so, you must account for that also!

  Your exchanger design is a mess. I suggest you start over. The more you try to fix the existing work (and you obviously put a lot of work into this), the behinder you will get. Try to find a worked example of an exchanger with steam condensing on the shell to use as a go-by. I am sympathetic as I made huge mistakes in academic projects. The goal is to learn if you can while also trying to make a good grade. After you calculate the flow rates and heat duty, then there are more tips you likely need for equipment design. Much of your calculations seem unnecessary and are probably adding to the confusion. 

 

Thank you for your input. Essentially I was trying to avoid phase change in the heat exchanger, by keeping the shellside above saturation temperature (6 bar), and tubeside below saturation temperature (7.6 bar), in order to stay out of the realm of phase-change calculations if possible. I had calculated the mass flow rate using my desired inputs/outputs of the heat balance equation, but perhaps that is not the best way to approach it. As you've pointed out, not allowing the shellside fluid to condense is resulting in a large pressure drop across my heat exchanger, which I'm assuming is to do with the fluid density affecting the mass velocity calculated from my mass flow rate. Instead, I will attempt to recalculate the heat duty and pressure drop assuming that the steam will condense on the shell-side. 

 

I have found limited results for open literature of shell-side condensation worked examples/calculations procedure. Coulsons chemical engineering book makes reference to "Engineering Sciences Data Unit Design Guide, ESDU 84023" for the procedure, however this is not available to me. I have found a thesis which appears to list the calculations for heat transfer coefficient and pressure drop for shell-side condensation, so I am planning to use this and try to validate individual equations from other sources.  (

 

Thank you very much

[charlie_KC]

Thank you very much for supplying those documents

Hi,

You may want to read this paper from one of our member .

Edit : a new resource added .

Good luck

Breizh

[Pilesar]

I made similar choices and mistakes in my senior design project and don't feel so alone now that you made the same mistake. You cannot keep steam superheated when used for heating in an exchanger. Steam in an exchanger does not condense because the bulk vapor gets colder, but because it contacts a cold metal surface. (You find the outside surface of your iced drink wet even though it is not raining in your house.) This is why the exchanger surface area is so important to heat transfer. Your spreadsheet showed a high heat transfer coefficient. That is only true while the steam condenses. Superheated steam has a very low heat transfer coefficient and is not suitable for use as a heating medium. When saturated steam condenses in an exchanger, the water takes up much less volume. The pressure inside the exchanger is reduced and more steam enters to take its place. Steam continually enters, contacts the surface area, and condensate is removed. Condensing heat exchangers are designed to help that process by keeping nozzles large enough to allow good flow and keeping a low pressure drop in the path of the steam to the tube surface. Baffles are minimized with large spacing and maximum cut -- really just enough baffles to support the tubes. The steam inlet may be near the middle of the shell so that there will be less pressure drop to the farthest exchanger surface. The tubes are arranged so that the condensate from one tube will not fall onto another tube surface (as much as possible) since the quicker the steam can contact the tube, the faster it can condense. Fancy tubes are offered that are not smooth on the outside which provide better sites for condensation to happen. These tubes are more expensive but can be useful to upgrade an existing exchanger installation when it may be an advantage to keep the same size shell.

[breizh]

Hi ,

Just an opinion , if you really want to learn about heat transfer 

better to get a copy of  : Process heat transfer by D.J Kern (version 1) or the updated version 2

or 

Heat transfer in process engineering by Eduardo Cao 

 

Note : You may want to consider this resource underneath :

           CheCalc ‐ Heat Transfer

 

Breizh 

[charlie_KC]

Would your recommendation be to convert the superheated steam to dry saturated steam before use in the heat exchanger then? eg. through the use of a de-superheater? The service fluid steam input currently specified in my work, has been bled from a turbine gland-sealing outlet line. 

 

My only issue is that the 'cold' process fluid (currently allocated to tubeside), requires an output of at least 150 degrees C. So if I reduce the temperature of my service fluid (steam) too much in order to achieve saturated steam, it may not be able to achieve the heat duty I require. Looking at a steam chart, it shows that the saturation temperature of dry steam is approx. 165 degrees C. 

I made similar choices and mistakes in my senior design project and don't feel so alone now that you made the same mistake. You cannot keep steam superheated when used for heating in an exchanger. Steam in an exchanger does not condense because the bulk vapor gets colder, but because it contacts a cold metal surface. (You find the outside surface of your iced drink wet even though it is not raining in your house.) This is why the exchanger surface area is so important to heat transfer. Your spreadsheet showed a high heat transfer coefficient. That is only true while the steam condenses. Superheated steam has a very low heat transfer coefficient and is not suitable for use as a heating medium. When saturated steam condenses in an exchanger, the water takes up much less volume. The pressure inside the exchanger is reduced and more steam enters to take its place. Steam continually enters, contacts the surface area, and condensate is removed. Condensing heat exchangers are designed to help that process by keeping nozzles large enough to allow good flow and keeping a low pressure drop in the path of the steam to the tube surface. Baffles are minimized with large spacing and maximum cut -- really just enough baffles to support the tubes. The steam inlet may be near the middle of the shell so that there will be less pressure drop to the farthest exchanger surface. The tubes are arranged so that the condensate from one tube will not fall onto another tube surface (as much as possible) since the quicker the steam can contact the tube, the faster it can condense. Fancy tubes are offered that are not smooth on the outside which provide better sites for condensation to happen. These tubes are more expensive but can be useful to upgrade an existing exchanger installation when it may be an advantage to keep the same size shell.

[Pilesar]

The superheat in your steam is great for turbines but will not help your exchanger since its heat transfer is so low. With superheated steam, the exchanger will need more surface area than if you had saturated steam at the same pressure. Some steam superheat is normal, but it is not helpful inside the exchanger. In real life, it is not normal to desuperheat the steam before the exchanger inlet due to cost and complexity of controls although it is sometimes done. Usually, the exchanger is designed with extra area to accommodate the superheat as long as the superheat is not too excessive. Your service has a small duty, so you would not add an external desuperheater just for this exchanger but you might for the whole steam header if your steam temperature is excessive and you have many other steam heating needs since adding water directly to the steam for desuperheating also generates more usable steam. The steam header still needs some superheat to keep dry. Your heat transfer will occur at the saturation temperature of the steam since it is the condensing that supplies the heat. Remember that the outside of the tubes will be mostly wet and will not see much superheated steam directly. The easiest calc and design would be to use steam at dewpoint conditions with condensate leaving at bubble point. That way you have to account for only the phase transition. This is often used in rough estimating but in real life the steam will usually have some superheat and the condensate will leave with some subcooling. Perhaps a good approach would be to size the exchanger assuming steam saturation conditions and then calculate how much additional surface area is needed for desuperheating (it may be substanial using the lower heat transfer coeffiicients) and include that additional surface area in the exchanger. This approach for your calcs will show you the relative surface areas for the desuperheating and condensing requirements compared to the duty transferred. Then you will see the advantage of saturated steam. For academic hand calcs, showing this two-step process might also demonstrate to the professor that you have a somewhat rigorous approach and are not just trying to change conditions to make the calcs simpler. Look hard for a go-by in your heat transfer references as I think it would help more than I can do.

[Bobby Strain]

You can still have wet tubes with superheated inlet steam . Looks like you are going to spend a lot of effort and time with this. Check your favorite search engine (more time & effort).

 

Bobby

[breizh]

Hi,

A good resource about steam and heat transfer :

https://www.spiraxsa...arn-about-steam

 

Enjoy the reading

Breizh

[breizh]

Hi,

You may want to learn about desuperheater .

Breizh 

[charlie_KC]

Thank you very much for providing that resource. Great explanations about the limitations and uses of superheated steam for a heat exchanger. This will help a lot. 

 

 

Hi,

A good resource about steam and heat transfer :

https://www.spiraxsa...arn-about-steam

 

Enjoy the reading

Breizh

[charlie_KC]

Hi both,

 

Just wanted to say thank you very much for helping me better understand the theory behind my mistakes. By reducing the amount of steam that is superheated, and allowing it to condense, I have been able to perform relatively successful preliminary calculations for my heat exchanger design. The pressure drop is well within limits, and it meets the required  heat duty. I will be talking about this learning curve within my dissertation. 

 

Kind regards

Charlotte

 

 

The superheat in your steam is great for turbines but will not help your exchanger since its heat transfer is so low. With superheated steam, the exchanger will need more surface area than if you had saturated steam at the same pressure. Some steam superheat is normal, but it is not helpful inside the exchanger. In real life, it is not normal to desuperheat the steam before the exchanger inlet due to cost and complexity of controls although it is sometimes done. Usually, the exchanger is designed with extra area to accommodate the superheat as long as the superheat is not too excessive. Your service has a small duty, so you would not add an external desuperheater just for this exchanger but you might for the whole steam header if your steam temperature is excessive and you have many other steam heating needs since adding water directly to the steam for desuperheating also generates more usable steam. The steam header still needs some superheat to keep dry. Your heat transfer will occur at the saturation temperature of the steam since it is the condensing that supplies the heat. Remember that the outside of the tubes will be mostly wet and will not see much superheated steam directly. The easiest calc and design would be to use steam at dewpoint conditions with condensate leaving at bubble point. That way you have to account for only the phase transition. This is often used in rough estimating but in real life the steam will usually have some superheat and the condensate will leave with some subcooling. Perhaps a good approach would be to size the exchanger assuming steam saturation conditions and then calculate how much additional surface area is needed for desuperheating (it may be substanial using the lower heat transfer coeffiicients) and include that additional surface area in the exchanger. This approach for your calcs will show you the relative surface areas for the desuperheating and condensing requirements compared to the duty transferred. Then you will see the advantage of saturated steam. For academic hand calcs, showing this two-step process might also demonstrate to the professor that you have a somewhat rigorous approach and are not just trying to change conditions to make the calcs simpler. Look hard for a go-by in your heat transfer references as I think it would help more than I can do.

 

 

Hi,

A good resource about steam and heat transfer :

https://www.spiraxsa...arn-about-steam

 

Enjoy the reading

 

[Art Montemayor]

This thread explains one of the reasons I've worked intermittently on transcribing the Spirax Sarco Tutorial on Steam in order to offer it to young students and engineers.  My academic experiences on heat transfer were parallel with those of Bobby Strain and Pilesar.  No one really taught or explained the simple ground rules of handling and dealing with steam out in industry until I met my first field mentor in Jamaica in 1961.  Alf Newton was an English engineer who learned his engineering in the British Navy, aboard steam-powered ships.  I was very lucky and fortunate when I came under his wings.  From him I learned all the real, hard basics of generating, transporting, handling, and using steam for heating and for power.

 

You need superheated steam for two reasons: to transport it and to generate power.  When you need steam for heat exchangers, superheat is a pain in the neck.  Pilesar has explained the basics of why.  Since my youth I have used a lot of superheated steam and every time I have had to use it in heat exchangers, I have always employed de-superheaters - as explained by Breizh.   You will find it very capital expensive to justify desuperheating in a heat exchanger.  Always use saturated steam to heat and rely on only transferring its latent heat.

 

Maintain the pressure of the saturated steam on your shell side and allow the created condensate to determine the latent heat transferred.

 

Study Kern for Process Heat Transfer - there is none better to date.  Coulson and all the dandies of today don't come close to Don Kern in teaching the fundamentals of heat transfer.  That's why his book is over 800 pages long.

 

The bad news is you've wasted a lot of your time in the calculations; the GOOD NEWS is that you've discovered and hopefully learned some very basic, fundamental, and powerful tools regarding the use of saturated and superheated steam.  Good luck.