Im doing a MDEA natural gas treating plant in ASPEN with bad results. Anyone who know how the solution to gas ratio and the solution composition should be?
It is difficult to tell unless we know the composition of a sour gas stream and what are the purity requirements. Can you be more specific about your problem? Like give some more details?
What property package are you using in ASPEN? The results are strongly dependant on that.
I have gained some more info now. Looking at BASFs homepage it says feed gas capacity in Nm3/hr = 2000 - 880000 and solvent circ.rate = in t/hr = 25 to 4500. In my calculations this makes roughly a L/G rate 7 - 18 kg/kg (calculating with pure CH4) Solvent concentration is between 35 to 50 % (molar I suppose)
The biggest issue in my simulation is that almost 1/4 of the CH4 follows with the solvent. In ASPEN Im using RADFRAC and the datapackage KEMDEA with 23 trays (sieve)70% flooding 13 bar. I noticed a lower followthrough when lowering the pressure but it never goes below 1/6 which is a too large fraction to loose. All inputs at top and bottom of ABS tower
I have worked with different simulations fro amines. In my experience, ASPEN has not powerful simulation for the same. HYSYS and TSWEET have special packages for amines developed by specialist in that area. Therefore they are more reliable.
Do you have these alternative simulations?
I am not sure if you will be able to set the VLE related parameters in ASPEN in this case as you are using tailor made package for MDEA. If you can do it, then check the parameters. I will also try to look in it.
Im afraid I only have ASPEN available. What is the VLE parameters? I will check with my boss if it could be available elsewhere. Is tsweet better than HYSYS?
Dear all I think ASPEN is ok if all thermodynamic properties of the components are available. However you may create a subroutine in e.g.Fortran for better reliable and accuracy results, but thats a long road to go !!
I gues you can retrieve the VLE parameters through results. I have some experience before with gas dehydration.I think by trial and error varying the values of solvent flow maintaining your spec you can arrive to an optimum point.
I believe wati is mistaken. In order for a simulator to properly simulate an amine absorption system, reaction kinetics must be taken into account. This process is not your typical absorption. HYSYS, TSWEET and others have the reaction kinetics of various amines with acid gases built in to the simulators (I believe HYSYS uses the reaction kinetics to come up with some sort of pseudo tray efficiency....this is activated when the amines package is used, which by the way costs extra money).
But those packages are available in ASPEN as well. I downloaded a package called KEMDEA. The simulation is done with the reaction kinetics the equlibrium modell is called ELECTRNL (I think) My guess is that its the same as HYSYS uses since ASPEN bought HYSYS a couple of years ago.
Although Aspen has bought HYSYS, HYSYS simulation package is as it was before. They are still working separately.
Three major simulation engine in this area are: TSWEET, HYSYS and ASPEN
ASPEN has some amine property packages which are based on some VLE theories (electrolyte-NRTL, UNIQUAC etc.). It is difficult to get the required parameters for these property packages as they are build in house and all the simulation companies dont want make them open
HYSYS has amine property package and amine simulation which was developed for them by DB Robinson group. They were sepcailized in that area and that is why their simulation engine was quite good at that time. But as that group is not existing now and hysys has not put much effort to further develop it, it is a little bit old simulation tool. However, it is still valuable and does the good job. As said earlier by Zoobie, it uses reaction kinetics to calculate the component efficiency. Both efficiency and equilibrium package are important to model these processes. In most of the simulation engines (aspen, hysys and tsweet), the model is fitted to experimental data and therefore gives reasonbale results. But one has to be careful with MDEA as the reaction chemistry is slightly different for this amine.
For TSWEET, it is still a active group and continuously improving their simulation tool for amine. Again they are expert in this area and most important is you can get help from them if you have any difficulty to model or need any specific advice. They have different models (equilibrium as well as kinetc with different thermodynamic packages )and vast experince to model these systems. I dont have specific experience with this tool. I attended on workshop arranged by them to introduce their simulation package. It is quite good tool.
BUT..every simulation engine has some advantages and disadvantages with them. However, HYSYS and especially TSWEET have remained quite popular with industry people.
There is another candidate in this area PROTREAT. It uses rate based modelling to model these processes. Principally, rate based models should give more accurate results. I dont have any experience about this simulation engine and therefore cant comment much on it.
This is just a small overview about the different simulation packages.
About your problem
1. I dont know the compositions and specific data about your system, but it seems that the property package is making some trouble in your case. If it is possible to give your data, i can just give a quick check using HYSYS. Just to check if it is a problem of using property package. you can send your data to my email address p.patil@manchester.ac.uk
2. In aspen, efficiencies are not calculated directly. Try to put the approximate efficincies for acid gases and see if this has any effect onlyour solution. But unfortunately, it is difficult to predict the efficiency as it is a complex function of temperature, pressure, hydrodynamics and composition also. But you can get some feeling about them. If you have a mixture of H2S and CO2, use 2 to 8 % for CO2 and 30 to 50 % for H2S on each stage. For other components you can assume to be in equilibrium i.e. 100 %
