Jump to content



Featured Articles

Check out the latest featured articles.

File Library

Check out the latest downloads available in the File Library.

New Article

Product Viscosity vs. Shear

Featured File

Vertical Tank Selection

New Blog Entry

Low Flow in Pipes- posted in Ankur's blog

5

Cryogenic Distillation And Cold Box


29 replies to this topic
Share this topic:
| More

#1 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 10 February 2021 - 05:46 AM

Hi there, 

 

As part of designing an ethylene and propylene production plant from steam cracking, I am currently working on designing the deethanizer, demethanizer and one of the refrigerant cycles (part of the cold box). We have selected a front end de-ethanizer (dual Pressure) with front end hydrogenation as our column sequencing order. 

 

 Another member of the group is working on the cracking furnace and so I don't have feed composition currently. I'm a little lost with where to start. We are required to complete the following (for the demethanizer)

  • Draw a PFD including control & relief streams - I assume this would just be columns +condenser and reboilers adjacent to it? - Is there extra equipment I'm forgetting? 
  • Mass and energy balances - Unsure on performing these? Given there is no reactions occuring just separations. 
  • Short cut - design to gain dimnesions - I feel the most suitable approach to shortcut sizing would be applying Fenske, Underwood, Gilliland and Kirkbride method - to calculate number of stages 
  • Application of thermodynamic models - I know that from using models e.g PRSV or SRK we could determine the relative volatitilites and understand the degree of separation. 

I just think I'm a bit lost with where to start. I would really appreciate some help and guidance getting started. 

 

 

Thanks 

 

Jai 



#2 SilverShaded

SilverShaded

    Veteran Member

  • Members
  • 43 posts

Posted 10 February 2021 - 06:22 AM

I'll give you my tuppence worth and others can disagree.

 

For the heat and mass balance the easiest, but not the only way, is to build a simulation and use that as the basis.  Also for sizing the column, I personally would not bother with short cut methods when you allready have a simulation, just use the the simulation.  Use a typical duty, reboil rate or reflux ratio that you know for a similar column and add/remove trays or reboil duty as needed to meet the product separation (or vary tray efficiency to do the same) then finalise with actual trays at a typical tray efficiency.

 

The optimium design of column is a trade off between energy costs for reboiling vs capital cost for adding trays or diameter.  So knowing the typical number of trays and reflux ratio in a similar existing column is a usefull guideline for a quick preliminary design, but does not gaurantee its an 'optimum'.

 

I would build a simulation, draw the PFD, add the Heat and Mass balance data to the PFD from the simulation.


Edited by SilverShaded, 10 February 2021 - 07:44 AM.


#3 Pilesar

Pilesar

    Gold Member

  • Members
  • 676 posts

Posted 10 February 2021 - 12:59 PM

Assume a reasonable cracking slate from the furnaces and begin. Expect to rework your design multiple times anyway. Your product specs will set your material balance for the most part since you have purity specs to meet. (Look backwards from the products to see where the impurities will come from and that will tell you what column specs are required.) Get your material balance completed before you start detailed design. Shortcut calculations can help determine duty requirements. Steam crackers use process fluids for heating and cooling where possible. Heat integration can be tricky so don't rush it. The common desire is to start designing equipment as soon as you can but spend extra time on the material balance and PFD. Don't expect your assignment to be easy... it will be more difficult than it first seems.



#4 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 11 February 2021 - 03:44 AM

Hi Silvershaded, 

 

I have determined the feed into the DeEthanizer and the demethanizer by assuming sharp splits so I have composition. I am looking for operating conditions currently (Temp, Pressure) . I have been told I need to start with mass balances for the demethanizer, then move onto the energy balances before designing. For a detailed mass balance, what software should I use? Excel seems like a good bet. In addition, I will need to make some assumptions and set key and non key components, and source relative volatilities etc. I am attaching a BFD of the units sequencing. 

I'll give you my tuppence worth and others can disagree.

 

For the heat and mass balance the easiest, but not the only way, is to build a simulation and use that as the basis.  Also for sizing the column, I personally would not bother with short cut methods when you allready have a simulation, just use the the simulation.  Use a typical duty, reboil rate or reflux ratio that you know for a similar column and add/remove trays or reboil duty as needed to meet the product separation (or vary tray efficiency to do the same) then finalise with actual trays at a typical tray efficiency.

 

The optimium design of column is a trade off between energy costs for reboiling vs capital cost for adding trays or diameter.  So knowing the typical number of trays and reflux ratio in a similar existing column is a usefull guideline for a quick preliminary design, but does not gaurantee its an 'optimum'.

 

I would build a simulation, draw the PFD, add the Heat and Mass balance data to the PFD from the simulation.

 

Attached Files



#5 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 11 February 2021 - 04:01 AM

I would agree that this will be a very challenging, its why I'm a bit stuck!! Haha- WIth working backwards I only know polymer grade ethylene needs to be produced (99.9%) which would be the output of the overhead stream from the C2 splitter. I have composition info coming out of the deethaniser but no information on acetylene hydrogenator or how it will remove the acetylene. For a preliminary mass balance how should I assume it behaves so I can gain composition data for the demethanizer? 

Assume a reasonable cracking slate from the furnaces and begin. Expect to rework your design multiple times anyway. Your product specs will set your material balance for the most part since you have purity specs to meet. (Look backwards from the products to see where the impurities will come from and that will tell you what column specs are required.) Get your material balance completed before you start detailed design. Shortcut calculations can help determine duty requirements. Steam crackers use process fluids for heating and cooling where possible. Heat integration can be tricky so don't rush it. The common desire is to start designing equipment as soon as you can but spend extra time on the material balance and PFD. Don't expect your assignment to be easy... it will be more difficult than it first seems.

 

 



#6 Pilesar

Pilesar

    Gold Member

  • Members
  • 676 posts

Posted 11 February 2021 - 08:07 AM

The basic reaction for acetylene hydrogenation is 'acetylene plus hydrogen produces ethylene.' The added hydrogen breaks the triple bond between the two carbons and leaves a double bond. This reaction is over catalyst and it is very efficient. If this equipment is in your scope of design, an internet search should give you some examples of yield.

For your material balance, use a computer. You will need to tweak it later so hand calcs are awkward. A spreadsheet will work to begin. Better is using a steady state process simulator if you have access. Each component separation will have impurities. You will need to set the impurities based on your product specs. The process will have anchor points for pressure and temperature to be consistent with the stream composition at the final product purity separation. For example, ethane in the propylene product comes from the bottom of your deethanizer. Thermodynamics will dictate the boiling point temperature required at the bottom of the deethanizer for the composition at the selected pressure. You must make sure the hot utility you select to reboil the deethanizer is appropriate. Because a steady state simulator has the thermodynamic methods included in the calculations, it should be more accurate than a spreadsheet and will give a more precise material balance.. Perhaps someone on your team has access to a simulator. Part of engineering is to find the resources you need. Check with your professor to confirm how much teamwork is allowed.

For senior design projects, I think the key is documentation. Show your work. Describe your methods. State your assumptions. Make notes for yourself while doing the work -- a journal is handy for when you need to backtrack and fix mistakes and also for when you write up your final report.



#7 ChEf

ChEf

    Junior Member

  • Members
  • 17 posts

Posted 12 February 2021 - 05:30 AM

I would suggest a modification to the scheme. The hydrogen needed for the acetylene conversion is already there in the feed stream to the front end acetylene converter.

The MAPD hydrogenation reactor, on the other hand, requires a hydrogen stream.  The demethanizer overhead, after partial condensation in the cold box, should be partially sent to a PSA unit, to recover the required hydrogen. Hydrogen could be used in some downstream polymerization processes, in addition to the MAPD hydrogenation reactor. The tail gas from the PSA Unit are usually sent to the furnaces to be used as fuel gas for the burners. Tail gas could also as regeneration gas for dehydration beds.



#8 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 15 February 2021 - 04:25 AM

I am looking for sources of operating temperatures, pressures and degrees of separation for de-ethanizer, depropanizer, demethanizer, C2 and C3 splitters  These are going to be used to form the preliminary mass balance. Ullmann's Encyclopedia has given some ranges for some of these units but we need some more specific temperatures and pressures so we can place coolers, condensers, heaters etc if necessary. 

 

 

Does anyone have any sources they could recommend? 



#9 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 15 February 2021 - 04:41 AM

We started with yields and selectivities in the furnace and determined its outputs and thus the feed into the cold section. We then to start with assumed sharp Splits across all separators. So now we are trying to find operating conditions and degree of separations to modify and improve these mass balances. I'm not too sure on the purpose of the cold box other than to cool the products from the cracking furnace down? Would the cold box be located before the front end de-ethanizer as well? 

 

Assume a reasonable cracking slate from the furnaces and begin. Expect to rework your design multiple times anyway. Your product specs will set your material balance for the most part since you have purity specs to meet. (Look backwards from the products to see where the impurities will come from and that will tell you what column specs are required.) Get your material balance completed before you start detailed design. Shortcut calculations can help determine duty requirements. Steam crackers use process fluids for heating and cooling where possible. Heat integration can be tricky so don't rush it. The common desire is to start designing equipment as soon as you can but spend extra time on the material balance and PFD. Don't expect your assignment to be easy... it will be more difficult than it first seems.



#10 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 15 February 2021 - 04:43 AM

Yes we are adding a hydrogen unit out of the demethanizer overheads to recover the hydrogen. We are debating the alternative technologies whether its a PSA or another option. Do you think PSA is sufficient? 

I would suggest a modification to the scheme. The hydrogen needed for the acetylene conversion is already there in the feed stream to the front end acetylene converter.

The MAPD hydrogenation reactor, on the other hand, requires a hydrogen stream.  The demethanizer overhead, after partial condensation in the cold box, should be partially sent to a PSA unit, to recover the required hydrogen. Hydrogen could be used in some downstream polymerization processes, in addition to the MAPD hydrogenation reactor. The tail gas from the PSA Unit are usually sent to the furnaces to be used as fuel gas for the burners. Tail gas could also as regeneration gas for dehydration beds.



#11 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 15 February 2021 - 05:05 AM

My notes from Sinnott's Chemical Engineering design are saying that to determine operating pressure: 

 

What feed type is most suitable for entering the column, saturated liquid, gas mixed feed etc?

 

"When selecting operating pressure its important  to ensure dew point of distillate is above the dew point of cooling water/fluid.

 

To calculate reboiler and condenser temperatures the bubble and dew points need to be calculated. Would doing this give the top and bottom temperatures from which I can calculate an average temperature of the column? 

Let me know your thoughts? 



#12 Pilesar

Pilesar

    Gold Member

  • Members
  • 676 posts

Posted 15 February 2021 - 12:11 PM

Distillation columns in operation are colder at the top than the bottom. The difference may be substantial as in the demethanizer. Average column temperature is not very useful to characterize a column. You should think through the process step by step to learn. Light hydrocarbons condense at colder temperatures. Methane will be the toughest to condense to provide reflux to the demethanizer. High pressure and low temperature are needed to condense methane. The overhead vapor of the demethanizer will therefore be very cold at high pressure. To keep from wasting the effort to make that stream so cold, the cold box allows the cold overhead stream to pre-cool the demethanizer feed. The demethanizer feed is cooled in stages. The liquid from each stage has a different composition so this allows optimizing their feed locations at the demethanizer. The cold box is designed as part of the demethanizer feed system. You can start demethanizer design based on assumptions, but won't be able to finish the demethanizer design until the demethanizer feed system is completed. Demeth feed system is one of the most complex design problems in the olefins plant. Since you have only the refrigerant scope of the cold box, you will have to get the final refrigerant specs from the demeth feed system designer. The refrig system won't be too hard to design once you understand how it works. The designer of the demeth feed system will be a bottleneck for your portion. I don't think you can wait to get started, so makes some assumptions and document them well to begin. (I've designed most all parts of new olefins plants except for the furnaces and worked in several plants but they were all demethanizer-first designs. Some of my comments above may not be completely relevant to your configuration.) Check with the provider of your simulation software to see if they can provide an olefins plant example.



#13 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 16 February 2021 - 04:46 AM

Thanks for this, Based on our research a front end deethanizer, appears to be a better approach, we have some sources justifying this but we can find no literature or case studies of this being applied. As you said most plants have designed front end demethanizer approaches. The interstage cooling between exchangers in the cold box is applicable but since we are using a deethanizer front approach, it is pre cooling the deethanizer feed. I'm also responsible for at least part of the cold box or de-eth feed system. So do you think we should start there first? 

 

I know the method to short cut design the column will be applying Fenske, Underwood, Erbar-Maddox, Kirkbride, which can help me then size the column. But what I'm unsure about are the mass and energy balances associated with column. I'm not asked to perform a stage by stage balances so I'm not sure what the balance would entail other than feed, distillate and bottoms compositions and flow rates? Since there is no reaction happening. 

Distillation columns in operation are colder at the top than the bottom. The difference may be substantial as in the demethanizer. Average column temperature is not very useful to characterize a column. You should think through the process step by step to learn. Light hydrocarbons condense at colder temperatures. Methane will be the toughest to condense to provide reflux to the demethanizer. High pressure and low temperature are needed to condense methane. The overhead vapor of the demethanizer will therefore be very cold at high pressure. To keep from wasting the effort to make that stream so cold, the cold box allows the cold overhead stream to pre-cool the demethanizer feed. The demethanizer feed is cooled in stages. The liquid from each stage has a different composition so this allows optimizing their feed locations at the demethanizer. The cold box is designed as part of the demethanizer feed system. You can start demethanizer design based on assumptions, but won't be able to finish the demethanizer design until the demethanizer feed system is completed. Demeth feed system is one of the most complex design problems in the olefins plant. Since you have only the refrigerant scope of the cold box, you will have to get the final refrigerant specs from the demeth feed system designer. The refrig system won't be too hard to design once you understand how it works. The designer of the demeth feed system will be a bottleneck for your portion. I don't think you can wait to get started, so makes some assumptions and document them well to begin. (I've designed most all parts of new olefins plants except for the furnaces and worked in several plants but they were all demethanizer-first designs. Some of my comments above may not be completely relevant to your configuration.) Check with the provider of your simulation software to see if they can provide an olefins plant example.



#14 breizh

breizh

    Gold Member

  • Forum Moderator
  • 4,942 posts

Posted 16 February 2021 - 05:58 AM

http://metsocontrols...ерегонка_en.pdf

 

Hi,

Using your favorite search engine . 

Good Luck

Breizh 



#15 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 16 February 2021 - 11:28 AM

I guess I'll keep looking thats what my lecturer said :) 

http://metsocontrols...ерегонка_en.pdf

 

Hi,

Using your favorite search engine . 

Good Luck

Breizh 



#16 Pilesar

Pilesar

    Gold Member

  • Members
  • 676 posts

Posted 16 February 2021 - 08:13 PM

The order of the separation columns varies for olefins plant designs as you discovered. Many years ago, I looked at studies trying to determine which scheme is best. The results I saw showed that there is just not much difference in efficiency between them. When you have a large-scale plant, a small difference in efficiency can be significant, though. Where the real opportunity for efficiency exists is in the furnace where making a profitable slate of products can be tweaked. You already decided your scheme for your student project, but here is a paper describing simulation and optimization for a different separation train arrangement:  http://www.uploadmb....p?id=1613524044 This is a thesis paper which is freely available for scholarly purposes. If you find this paper useful, there are other thesis papers out there for olefins plants if you can find them.



#17 Pilesar

Pilesar

    Gold Member

  • Members
  • 676 posts

Posted 16 February 2021 - 08:29 PM

A mass and energy balance for a column will show the feed and product streams conveniently displayed in a spreadsheet showing composition, temperature, pressure, flow rate, and assorted other parameters that might be useful to an equipment designer for each stream. These parameters would include molecular weight, density, surface tension, viscosity, etc. The associated exchanger duties (typically condenser and reboiler) might all be located on separate spreadsheet page with a utility summary indicating steam requirements, cooling water requirements, etc. The streams in the balance will be referenced to a process flow diagram showing their location in the process.

A distillation stage by stage balance is not part of the overall material balance deliverable, but it will be required for designing column internals in a similar manner that heating curves are required for exchanger design.



#18 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 17 February 2021 - 06:05 AM

Thanks for this, most papers I have read from science are demethanizer first approaches and this is the challenge. As we selected a de-ethanizer first approach as this can process a feed w/ large perefcentage propane, results in less fouling, and in general has lower operating costs. Compared with a front end depropaniser it has less equipment reducing CAPEX costs. Its good this paper is a front de-propanizer, so I'm sure I will find a deethanizer approach. I just need to keep digging. 

The order of the separation columns varies for olefins plant designs as you discovered. Many years ago, I looked at studies trying to determine which scheme is best. The results I saw showed that there is just not much difference in efficiency between them. When you have a large-scale plant, a small difference in efficiency can be significant, though. Where the real opportunity for efficiency exists is in the furnace where making a profitable slate of products can be tweaked. You already decided your scheme for your student project, but here is a paper describing simulation and optimization for a different separation train arrangement:  http://www.uploadmb....p?id=1613524044 This is a thesis paper which is freely available for scholarly purposes. If you find this paper useful, there are other thesis papers out there for olefins plants if you can find them.



#19 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 17 February 2021 - 06:07 AM

We do not need to perform stage by stage analysis, we can simply select column internals based on recommendations from literature. But I will need to perform a slightly deeper mass balance (calcualte liquid and vapour flowrates in rectifying and strippings section) but that should be ok. I'm not too sure on the energy balance, other than the duties of the condensers/reboiler. I will need to determine enthalpies of all the streams in and out. 

A mass and energy balance for a column will show the feed and product streams conveniently displayed in a spreadsheet showing composition, temperature, pressure, flow rate, and assorted other parameters that might be useful to an equipment designer for each stream. These parameters would include molecular weight, density, surface tension, viscosity, etc. The associated exchanger duties (typically condenser and reboiler) might all be located on separate spreadsheet page with a utility summary indicating steam requirements, cooling water requirements, etc. The streams in the balance will be referenced to a process flow diagram showing their location in the process.

A distillation stage by stage balance is not part of the overall material balance deliverable, but it will be required for designing column internals in a similar manner that heating curves are required for exchanger design.



#20 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 17 February 2021 - 12:18 PM

An interesting point made, We have identified papers that gave degree of purity achieved and operating conditions needed for the overall mass balance. But the papers were for front end demethanizer approaches.  Our supervisor said using these operating conditions for our deethanizer approach would not be suitable. So I am curious as to why this is the case. Ullmann's encylcopedia gives a range but is not a strong enough reference to just pick within the range of temperatures and pressures. 

 

Linde AG's technology appears to be applicable to our plant and I've asked them for some data, so hopefully they respond. 

The order of the separation columns varies for olefins plant designs as you discovered. Many years ago, I looked at studies trying to determine which scheme is best. The results I saw showed that there is just not much difference in efficiency between them. When you have a large-scale plant, a small difference in efficiency can be significant, though. Where the real opportunity for efficiency exists is in the furnace where making a profitable slate of products can be tweaked. You already decided your scheme for your student project, but here is a paper describing simulation and optimization for a different separation train arrangement:  http://www.uploadmb....p?id=1613524044 This is a thesis paper which is freely available for scholarly purposes. If you find this paper useful, there are other thesis papers out there for olefins plants if you can find them.



#21 Pilesar

Pilesar

    Gold Member

  • Members
  • 676 posts

Posted 17 February 2021 - 03:01 PM

You want to operate distillation columns at as low a pressure as possible for the most efficient component separation. You want to operate at a pressure where you can condense and reboil with reasonable utilities - preferably avoiding refrigeration systems. You want to avoid expensive compressors where possible. Lower pressure distillation needs larger diameter columns to handle the lower density vapor flow. Higher pressure distillation has higher temperature at the bottom of the column which may lead to product degradation. Higher pressure equipment costs more to build for the same volume. Capital cost and energy usage are both important to plant economics. Capital costs and energy costs vary from year to year and region to region. Equipment availability, lead time, vendor support, site constraints, licensing costs all should be considered when designing a plant. Design decisions try to balance competing interests within constraints. The optimal decision for a plant in one site and time may be different from another plant. That most ethylene plants were not built using the scheme you selected does not mean that those designers were wrong -- they may have had different economics and constraints. For your project, you have a schedule to meet which limits your ability to optimize. Optimization is admirable, but less than optimal plant designs are the general rule in the real world. Good luck finding the references you want.



#22 PingPong

PingPong

    Gold Member

  • Members
  • 1,353 posts

Posted 19 February 2021 - 06:50 AM

An ethylene plant is probably the most complicated design in the petrochemical or petroleum industry. Reputable licensors like Lummus, Technip, Linde, KBR, spent many many years on optimizing their process schemes and furnace designs. I have worked on several designs myself in the past decades so I know that it's an illusion to think that a student can invent, or event select, an optimal process scheme within a few months, without having any prior design experience of even simple processes.

 

The most economical process scheme also depends on the feedstock of the plant. Sometimes a licensor will use a different process scheme for cracking gas (ethane, propane) than for cracking liquid (naphtha, gasoil, hydrowax). The feedstock also impacts the choice of cracking coil layout and its outlet temperature.

 

And of course there is often client interference: some client engineers will insist on this, or that, or such, or so, because of bad experience in the past (which could also have been due to bad plant operation), or simply because they think they know it better than the experienced engineers of the licensor. In your case you don't have client interference, but maybe your supervisor is trying to steer you into his preferred direction.

The best a student can do is copy a process scheme from a licensor, or from an existing plant.

Choose a simple scheme, otherwise you will certainly get overwhelmed and run out of time.

 

When choosing the pressure of any column one has to consider the resulting temperatures in condenser and reboiler, and how to achieve those.

 

The demethanizer (DeC1) is part of the cold section that also includes the chilling train (cold boxes, expander(s), et cetera) in which the multiple feeds to the DeC1 are condensed, as well as refrigeration systems. This whole system has to be viewed as one section. One cannot design a DeC1 without simultaneously designing the chilling train.

There are DeC1's operating at pressures ranging from as low as 7 bar to almost 30 bar. Anything is possible provided the rest of the cold section can supply the required conditions for feeds, condenser and reboiler(s).

 

The deethanizer (DeC2) system pressure(s) is chosen such that bottom temperature(s) does not exceed 80 oC to limit fouling (but still a spare reboiler is usually installed to allow online cleaning) and its distillate is dew point vapor with the correct maximum C3 content. You need to think about what refrigerant you will use, and at what pressure, in the condenser(s).

I strongly suggest you revise your BFD. Focus on what licensors do, not what amateurs do in a thesis or whatever.

See for example:

Attached Files



#23 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 20 February 2021 - 12:21 PM

We are following technip and Linde's front end deethanizer approach. But now having identified operating conditions, we have assumed sharp split in our ball park mass balances. But we need to find separation effficiencies of each column for the deethanizer approach. Most papers and literature we find provide separation efficiencies for front end demethanizer approaches. None found so far are for deethanizer approaches. We know that sharp split is not a great assumption and would prefer to modify the mass and eenrgy balances.  



#24 JaiEdi

JaiEdi

    Junior Member

  • Members
  • 23 posts

Posted 20 February 2021 - 12:39 PM

I also contacted Linde to ask for distillation separation data. No response. 



#25 Bobby Strain

Bobby Strain

    Gold Member

  • Members
  • 2,961 posts

Posted 20 February 2021 - 03:47 PM

The licensors won't respond to your request. You can easily determine the required separation from the product specifications. It's a shame that you must waste so much time on this rather that learning fundamentals.

 

Bobby






Similar Topics