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

0

Isothermal Water Ga Sshift-Simulation And Design


24 replies to this topic
Share this topic:
| More

#1 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 16 November 2021 - 05:26 AM

Hi, need help with isothemral water gas shift which is different from high temo or low temp shift.

 

I wanted to knowwhether it can be modeled as RGIBBS (or RSTOIC?) for the latter you need to know conversion beforrehand.

 

does anybody have any links/info on how typical iso shift reactors work(Linde, Lurgi etcZ) and what is preferred steam/CO ratio at inlet and typical per passconversion of CO? how close is it to equilibrium?

 

Just to add, these are used in hydrogen productrion plants for methanol and ammonia.

 

I am designing a reactor starting from simulation for one such hydrogen production plant of worldscale capacity( 70 KTPY).

 

The isothermal temperature I have seen in literature is around 250-260C for these reactors.

 

Please let me know the simulation approach fpor design. the inlet feed is coming from a primary reformer and has steam in excess, almost 4:1, steam : CO ratio on molar basis.

 

Pressure is around 22 bar



#2 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 19 November 2021 - 08:38 AM

An isothermal WGS reactor is merely a reactor with cooling tubes in it to remove the exothermal heat of reaction by generating steam (or whatever).

 

You can use them for high or medium or low temperature shift, depending on the type of catalyst you put into it.

In that respect it is not different from a conventional adiabatic reactor.

The molecules don't care about what kind of reactor it is, all they notice is temperature, pressure and catalyst.



#3 Pilesar

Pilesar

    Gold Member

  • Members
  • 1,347 posts

Posted 20 November 2021 - 04:48 AM

The shift and methanation reactions have well-defined stoichiometry and the conversion is essentially at equilibrium for the reaction temperature. Because just a few of the components take part in the reaction, i would choose REquil instead of RGibbs. It is possible that the results would be the same, but RGibbs may include reactions that you would prefer to be excluded. With REquil, you have better control over the resulting composition.



#4 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 20 November 2021 - 05:30 AM

Thanks Ping Pong/Piledsar.

 

Pilesar, industrially, do reactors approach equilibrium conditions? or should i use aproach to tempersature of say +10C to be closer to reality?

 

also in water gasshift reactors, do methanol formation happen? these reactors use copper based catalyst



#5 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 20 November 2021 - 05:33 AM

reactors are generally kinetically controlled in industrial operations, thats why i was asking, almost close to equilibrium but not exactlky at equilibrium, right?



#6 Pilesar

Pilesar

    Gold Member

  • Members
  • 1,347 posts

Posted 20 November 2021 - 09:33 AM

It's been three decades since I designed ammonia plants and details in my mind have gotten overwritten by later data input. For a student project, I doubt you need to put much approach, if any, to the shift reactions. If you want to put an approach in the model to let the professor know you understand the concept, I would suggest making it a very small one (maybe -5 C?). Note that the approach would be negative. Your literature search should find some documentation of how others model this. I am not sure what you mean by 'kinetically controlled.' Industrial operations want these conversions to equilibrium as much as possible. I tracked these reactors when working in an ammonia plant to monitor catalyst activity and I calculated the approach to equilibrium for all the reactions as part of my responsibilities.  I remember my primary and secondary reformers had approaches in the general vicinity of -40 F. The shift and methanation reactor approaches were much smaller than that. The calcs never showed exactly zero approach and I remember on occasion I would even see a slight positive approach which just indicated to me the uncertainty of relying on field data.


Edited by Pilesar, 22 November 2021 - 12:55 AM.


#7 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 20 November 2021 - 02:40 PM

If you want to put an approach in the model to let the professor know you understand the concept, I would suggest making it a very small one (maybe -5 C?). Note that the approach would be negative.

That is not correct.

 

In any case for WGS the ATE (approach to equilibrium) has to be added as a positive number to the actual reactor outlet temperature,

but moreover the ATE is always to be specified as a positive number because it is applied by the process simulator as follows:

 

For exothermic reactions (like WGS) the equilibrium constant Keq is calculated at T = Treaction + ATE


For endothermic reactions the equilibrium constant Keq is calculated at T = Treaction - ATE

 

For WGS an ATE of 5 to 10 oC is customary for design calculations.
 



#8 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 21 November 2021 - 05:25 PM

Pingpong, if i model SMR as RGIBBS say, reformer has both endothermic(reforming) and exothermic(shift) reactions happening at the same time. so how do you decide approach to temperature here? since the overall reformer is endothermic, do you just say mention it as -10C? as ATE for SMR?



#9 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 22 November 2021 - 08:50 AM

I don't know what simulator you are using, but in any case I advice you to study your reference manual and keyword manual.

Could be that there is special option for methanation (reverse of reforming) and shift reactors in it.

Verify also how your specific simulator applies the user entered ATE.

 

In PRO/II (which I use) there is option to select an equilibrium reactor that simulates a methanator. That covers both the shift and methanation (reverse of reforming) equilibria. Because shift and methanation are both exothermic a user specified ATE is applied to both in the same way. ATE to be a positive number.

 

How reforming/methanation can be modeled in your simulator I can't tell, check your manuals.


Edited by PingPong, 22 November 2021 - 09:12 AM.


#10 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 22 November 2021 - 09:52 AM

Pingpong, I use aspen plus which uses RGIBBS model block that uses Gibbs minimization of free energy. you dont need to mention reactions explicitly. This is generally good in situations where yopu dont know all the reactions that are going on inside the reactor, but know that the reactor is operating close to equilibrium.



#11 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 23 November 2021 - 08:18 AM

I know what a Gibbs reactor model is as PRO/II also has it, but I never felt the need to use it. I always use an equilibrium reactor and apply suitable ATE's.

 

What I wrote earlier about how a simulator like PRO/II applies user specified ATE in an equilibrium reactor will however not apply to a Gibbs reactor.

 

Not sure how it works in AspenPlus but I suspect that you can only specify one approach (ATE) which is then added to the temperature that is used for the Gibbs free energy minimization calculation. That's not ideal for simulating a steam reformer.

 

It should however be possible to specify the relevant reactions in RGibbs and apply an ATE to those, but you would need to study the AspenPlus reference guide how to apply that.

 

An old AspenPlus 10 reference guide that I found on the internet states:

"For temperature approach specifications, RGibbs evaluates the chemical
equilibrium constant at T + ΔT, where T is the actual reactor temperature
(specified or calculated) and ΔT is the desired temperature approach.


You can have RGibbs consider only a specific set of reactions. You can restrict the
chemical equilibrium by specifying temperature approach or molar extent for the
reactions. You must specify the stoichiometric coefficients for a complete set of
linearly independent chemical reactions, even if only one reaction is restricted."

 

The latest AspenPlus reference guide will be more accurate and useful.


Edited by PingPong, 23 November 2021 - 08:26 AM.


#12 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 28 November 2021 - 01:51 PM

Ping pong so is it enough if i define the 2 reactions as methane reforming and water ga sshift in aspen an ddefine appropriate ATE for both? one will be positive, other negativre.  and will you use these ATE as tunebale to achieve final target composition?

 

i wanted to know what are some realistic ASTEs you have used for SMR?



#13 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 03 December 2021 - 10:03 AM

Once again: I don't know how it is best done in AspenPlus as I don't use that simulator.

 

To get your simulation started and debugged you should not worry about what ATE to use, just use zero to start with.

After it all works correctly you can apply ATE.

For the reforming reaction the Tequil should be roughly 10 oC lower than the COT, for the WGS reaction Tequil should be roughly 10 oC higher than the COT.

If you have operating data then you adjust the ATE's until reformer effluent composition matches your actual data.



#14 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 03 December 2021 - 04:23 PM

PingPong, thanks, do you knowe the typical delta T, the temperature increase across a reformer with S/C close to 3? is 300C realistic?

 

and what is the typical methane conversion in the reformer? is it 70-80%?



#15 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 05 December 2021 - 02:52 PM   Best Answer

In designs from several licensors that I have seen in the past the feed gas (NG + steam) temperature entering the radiant coil is usually in the range of 550 to 650 oC.

Exact temperature depends on the C2+ content of the gas mixture, the catalyst used in the radiant coil, and the conservatism of the licensor. If a Prereformer is used there is no C2+ present so it could be close to 650 oC.

 

COT is usually in the range of 870 to 930 oC depending on the maximum methane slip one wants to allow, on the catalyst used in the radiant coil, and the conservatism of the licensor.

 

The methane slip is the volume percentage (mole percentage) methane that is present, on a dry basis, in the reformer effluent.

That is usually in the range of 3.5 to 5 v%.



#16 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 06 December 2021 - 08:29 AM

Pingpong, can you also provide a quick guideline on how to preform a preliminary design of the steam drum and the exchangers illustrated inside a SMR(where preheatinf of feed and steam happens via hot flue gas)? I'll break it.

 

1. can i model and design steam drum as just a flash vessel in simulator where really hot water or steam goes in and after phase separation saturated steam emerges which can be further heated or used in process? any articlesor links on design considerations for steam drum will be helpful

 

2. the coils inside the SMR where the heat transfer happens to preheat feed, steam etc. how can they be modeled/designed in simulator? they arent exactly shell and tube. what typre of exchanger should they be modeled as? simulator provides shortcut options as well as detailed options for these exchangers. 

 

Thanks in advance



#17 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 07 December 2021 - 05:51 AM

I strongly suggest you go through this topic in which I already explained many things related to SMR calculations: https://www.cheresou...-reformer-unit/

 

1) the reformer effluent at 870 - 930 oC is cooled in a steam generator down to roughly 350 oC (or whatever temperature suits your design best). Above it is a steam drum. The two together are a special item designed by a specialized company. In your simulation you simply use a Flash or Heater (or whatever suitable module in AspenPlus) to achieve this cooling and that provides you with a duty. Depending on the temperature of the BFW entering, and the amount of blowdown to be used, you can simply calculate the amount of steam produced by that duty by hand (using a steam table) or by making a spreadsheet for such simple calc.

 

2) see topic mentioned above.


Edited by PingPong, 07 December 2021 - 05:52 AM.


#18 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 08 December 2021 - 12:09 AM

pingping, that thread was a bit tedious to follow with all that hand calculations. I am using a simulator. I have a quick question. Lets say my NG+steam mixture enters SMR reactor at 550 and leaves at 860C. when i calculate fuel that needs to be burnt, which I am modeling as a combustion reactor starting with NG and air, the temperature of flue gas from this rreactor can be 1000C, even AFTER delivering the duty needed for the SMR reaction, right? i.e if I remove or subtract duty needed to be provided for the SMR reactor from the combustion product mixture, its net temperature is still 150C higher than SMR exit temperature, right?



#19 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 08 December 2021 - 07:20 AM

pingping, that thread was a bit tedious to follow with all that hand calculations. I am using a simulator.

In that thread you should only reads my posts with useful advice and comments. Ignore all other posts.

 

What is the purpose of your simulations?

Are you trying to model an existing SMR unit?

Or are you doing all this just for the fun of it?



#20 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 08 December 2021 - 08:24 AM

pingpong, I am trying to simulate and design an ammonia process for high level planning/CAPEX purposes. This SMR is part of it. I am getting some design info from open literature



#21 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 08 December 2021 - 08:25 AM

is there a guideline for how high the bridgewall temperature or temperature of gas leaving the radiant section after transferring the heat to reformer tubes can be? is this related to reformer exit temperature? (which is 860 in my case?)



#22 zuhroni130

zuhroni130

    Brand New Member

  • Members
  • 2 posts

Posted 13 December 2021 - 07:32 PM

pingping, that thread was a bit tedious to follow with all that hand calculations. I am using a simulator. I have a quick question. Lets say my NG+steam mixture enters SMR reactor at 550 and leaves at 860C. when i calculate fuel that needs to be burnt, which I am modeling as a combustion reactor starting with NG and air, the temperature of flue gas from this rreactor can be 1000C, even AFTER delivering the duty needed for the SMR reaction, right? i.e if I remove or subtract duty needed to be provided for the SMR reactor from the combustion product mixture, its net temperature is still 150C higher than SMR exit temperature, right?

daraj, so the purpose of your simulation is that you want to know the temperature of the fuel gas that is used in radiant section of SMR and coming out through the stack? Is that what do you want to know?



#23 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 14 December 2021 - 07:48 AM

pingpong, I am trying to simulate and design an ammonia process for high level planning/CAPEX purposes. This SMR is part of it. I am getting some design info from open literature

You should realize that designing such units is not for amateurs that depend on open literature or internet sources like this forum, unless it is merely for a study exercise or thesis.

 

is there a guideline for how high the bridgewall temperature or temperature of gas leaving the radiant section after transferring the heat to reformer tubes can be? is this related to reformer exit temperature? (which is 860 in my case?)

Usually the bridgewall temperature in an SMR furnace is roughly 150 oC higher than COT.
 


Edited by PingPong, 14 December 2021 - 07:50 AM.


#24 daraj

daraj

    Gold Member

  • Members
  • 221 posts

Posted 26 December 2021 - 04:13 PM

PINGPONG, For sizing the isothermal shift need to know what is the typical GHSV i can assume? using this i can find bed volume and then add some percentage to it to get reactor vessel volume, thanks



#25 PingPong

PingPong

    Gold Member

  • Members
  • 1,446 posts

Posted 27 December 2021 - 10:01 AM

In the earlier mentioned topic I gave some indication for HTS,

but I don't have such data for MTS in an isothermal reactor.






Similar Topics