Sie sind auf Seite 1von 52
Page 1
Page 1

Aspen Tech Handbook:

A Technical Aide for Chemical Engineering Process Design Students

Contents

Contents For information on the contents of any section, read the opening paragraph of that section,

For information on the contents of any section, read the opening paragraph of that section, or use the alphabetical index in the back of the handbook for easier searching.

Chapter

Title

Page

1 Computer Management

 

2

2 Multicomponent Flash Drums

6

3 Generating Graphs in Aspen

10

4 Modeling Distillation Columns

13

5 Modeling Reactors

20

6 Using RateFrac for Rate-Based Simulations

22

7 Putting Everything Together

 

26

8 Using Aspen Icarus for your Cost Estimate

34

9 Using Aspen Pinch For Your Heat Integration

39

Appendix A

Thermodynamic Property Tree

47

Appendix B

References

48

Appendix C

Index

49

Note: The makers of the AspenTech, Inc.’s Aspen Plus simulation engine have no affiliation with this tutorial. This tutorial was produced for educational purposes and any distribution of this material without the expressed written consent of Cornell University is strictly prohibited. Possible side effects of using this software include: blurred vision, self-induced hair loss, hours of frustration, and the realization that you probably never need to use activity coefficients again. Please consult your physician before operating the simulation engine.

Page 3

Preface to

the Manual

Over the last several years, we have introduced modern Chemical Engineering software to the Unit Operations and Design Courses. Even more recently, we have introduced AspenPlus simulation to the Separations Course and we hope this "downward" integration will continue. The value to the senior courses has been enormous and many graduates have expressed that knowing something about these computer techniques has been helpful in their jobs.

Unfortunately, as the programs get more able to predict real processes and equipment, they become more complicated. The "Help" manuals are not easy (although they are complete) and step-by-step tutorials are non-existent. Keep in mind that we are being provided with several hundred thousand dollars worth of software and licenses for a tiny fraction of the normal price. While Aspen does provide technical service, we cannot expect them to answer our specific questions overnight.

In the spring of '01, Brock Tuczynski was a TA for Separations and wrote a very nice tutorial for getting started with Aspen. I had written several sets of instructions for solving various problems with Aspen and Icarus and, over the years, we had assembled lots of specific but unorganized answers to questions.

Over the summer of '03, Josh Banke and Meghan Cuddihy put together this tutorial. They edited Brock's original work, tied together all we had and wrote many new sections. They have tried to make this into a comprehensive user's guide to the Aspen and Icarus software and have added quite a bit of material about the computer lab organization and setup that I hope you will find useful.

Good luck as you become more proficient with the simulations and remember that these are difficult and complicated programs to use. Patience is important. One last piece of advice: the fundamentals of Chemical Engineering are the most important tools you have. I have seen many, many flash drums that were not adiabatic, columns with a reflux ratio below the minimum( try a McCabe-Thiele Diagram for starters), and overall material balances that were not right. Start every problem with a sketch, a material balance and a heat balance and you will spend a lot less time trying to get yo ur process to converge.

Ken Ackley

Page 4

Chapter 1: Computer Management

The purpose of this chapter is mainly to demonstrate how to setup a shared folder on the computer lab network, how to optionally set up one from your home, as well as to discuss some of the more pertinent issues regardingmanaging and securing the integrity of your computer that are useful to know.

Windows XP File Sharing Windows XP lets you share a computer's disks and folders with other computers on the network, using a method called Simple File Sharing . If a disk or folder is shared, everyone with whom you give permissions to the file may access it from the network. This can be useful for several reasons: a shared folder will make it easy for all members of the group to access any saved Process Design or UO files; it obviates a need for writing files to discs or other data transfer media; and it can ward off any answer predators, who might want access to your Aspen or Icarus files.

Probably the simplest method for going about this is to share and set permissions on a folder in your account on the Instruct Domain. Here is a brief explanation of how to do that. o Note: The computers in the computer lab in Olin Hall should have Windows XP Professional, if not the latest operating system, installed and running, so we will try to base our example on that assumption. In our example, we've used Windows Explorer to browse to the directory of the My Documents folder, which should be located on the desktop.

o

o

o

o

o

o

o

o

o

In the right-hand pane, right-click, select New and then Folder, and enter the name Shared Folder (or whatever you feel comfortable naming it). o Note: For maximum compatibility with all versions of Windows, use 1-12 characters.

Now, you should specify sharing options for the folder you have created. To do this, right- click the folder and select Sharing and Security.

On the Sharing tab, select Share this folder and enter a share name. You can add a comment that describes the share on other computers on the same network.

Leave the User limit alone. On XP professional, the maximum limit is 10.

You should probably set some password permissions, so that only you or your group members can access it. To do this:

Click Permissions. Notice that, by default, the group Everyone has Full Control. This means that all users can read, write, and even delete files. That's not what we want at all!

Click Add, and then choose Object Types. Un-check Built-in security principles and Groups, because we only want to see Users. Click OK.

Choose Advanced, and click Find Now. Click on the users who should have access to this share. This should include your group partners and you. You can repeat this to add additional users. When you are done, click OK.

You should be back at the Permissions List. By default, the newly-added users have read-only access. You may want them to have read/write access, meaning each user who logs in can alter files as well as read them, so tick the Change box. You should repeat this for each user as you see fit.

A powerful alternative is to set up a share at home, if you don’t want to worry about being in the lab to access the shared folder. This way, you and your partners can telecommute and can exchange ideas and work without being physically present. You would just need to drop your work on the shared folder and it would be accessible by all members of your group. But this method has some issues with compatibility that need to be taken into account first.

with compatibility that need to be taken into account first. Important! Making Sure You Can Set
with compatibility that need to be taken into account first. Important! Making Sure You Can Set

Important! Making Sure You Can Set Up a Shared Folder at Home The process for setting up a shared folder that is accessible from another connection is not necessarily easy. There are a few requirements that need to be met first.

Before setting up a shared folder that can be easily accessed from home, either you need to be on the Resnet Domain, or you should have a direct connection to the internet with what is called a static IP given to you by your internet service provider. Being on Resnet automatically allows you to access your computer easily from home, but if you live off campus, and want to set up a share that you can access from outside home, you will need to make sure you have a static IP.

Your IP address is essentially the identifying address that your computer uses to interact with the internet, and, as the name suggests, a static IP address is one that doesn’t change. If you are not sure whether you have a static IP or its counterpart, a dynamic DHCP-based address, there is a way to check.

Page 5

o

o

o

First you need to know what your IP is, if you don’t already know it. There are a number of ways to go about

this. One way is to hit the Start button, then click the Run button. Type in “ cmd” and hit enter. This should bring up what is called a command prompt or a dos prompt.

At the prompt, type in “ipconfig,” and the result should be

a listing of numbers, but the one you will need to actually

take note of will be the number labeled by “IP Address.” Keep this prompt open. To make IP address changes from within a script, you can use Win2K's multipurpose Net Shell (Netsh) command. This command provides several functions that relate to viewing and changing IP addressing on a Win2K system. For example, to switch from DHCP to static IP, enter the following command:

netsh interface ip set address "[connection name]" static [ip address] [netmask] [gateway] [metric]

Netmask is the subnet mask (e.g. 255.255.255.0) associated with the IP address, gateway is the default gateway on the command prompt, and metric is the numerical difference between the IP and the default gateway. For example, to change to the static IP address 192.168.0.4, mask 255.255.255.0, and gateway 192.168.0.1 (metric 3), you would type:

and gateway 192.168.0.1 (metric 3), you would type: netsh interface ip set address "local area

netsh interface ip set address "local area connection" static 192.168.0.4 255.255.255.0 192.168.0.1 3

o

If you still have access to the internet, then your ISP supplied you with a static IP but if you no longer have access to the internet, then you

don’t have a static IP. To switch back to DHCP-based addressing, you need to merely enter into the command prompt:

netsh interface ip set address "<connection name>" dhcp

If you do not have Resnet or a static IP it isn’t the end of the world. You have 3-4 members per group, so there is a good chance that someone in the group has a direct internet connection with a static IP. But let’s be optimistic, and assume your apartment has viable access.

Windows XP Home Edition’s Simple File Sharing Windows XP lets you share a computer's disks and folders with other computers on the network, using a method called Simple File Sharing, and it really is simple. If a disk or folder is shared, everyone on the network can access it. There are no user permissions and no passwords. Because sharing in this way is so wide open, Windows XP tries to protect you from some potential security risks. Just a small note: Windows XP Professional has a much more powerful way to control the sharing of files, so if you have it or have access to it, I recommend it over the Home Edition. It’s a little more complicated, but you will worry less about who is rummaging through your

computer and who isn’t. However, since your computer more likely has the home edition installed on it, here is a short how-to for setting up a shared folder with XP Home.

o Right click the disk or folder that you want to share and select Sharing and Security. The disk or folder that you share, along with all of the folders that it contains, will be accessible by other network users.

folder that you share, along with all of the folders that it contains, will be accessible
folder that you share, along with all of the folders that it contains, will be accessible

Page 6

o Note: If you're sharing an entire disk, Windows XP gives a warning. What’s implied in the warning is that it's better to share a specific

folder, since only that folder (and its subfolders) will be accessible by others, and the rest of the disk will be inaccessible. Click where indicated if you want to go ahead and share the entire disk. This screen doesn't appear if you're sharing a folder.

o

The first time that you set up sharing, Windows XP displays a warning, urging you to use the Network Setup Wizard for safety. Click where indicated to either run the Wizard or to share files without running the wizard.

o

Note: The Wizard automatically enables the Internet Connection Firewall (ICF) to prevent other Internet users from accessing your shared disks and folders. Enabling ICF is a good idea if you connect directly to the Internet through a dial-up, DSL, or cable modem. But it's a terrible idea if you connect to the Internet through your LAN, using a software or hardware router, since it will block File and Printer Sharing.

o

XP should display another warning. If you want ICF enabled, select Use the wizard to enable file sharing. Otherwise, select Just enable file sharing.

 
 
 

o

Having successfully run the Wizard's obstacle course, you may now specify a Share name, which users on other networked computers will use to access this disk or folder.

o

By default, users on other computers have full access: they can read, write, and delete shared files. If you only want them to be able to read files, un-check Allow network users to change my files.

o

Warning: If a user has full access, deleting a file doesn't put it in the Recycle Bin. Once it's deleted, it's gone for good. Also, if you don’t have an Antivirus program, there is a good chance to pick up a virus accidentally from one of the computers in the lab. We strongly urge you to never run Aspen without an updated anti-virus agent, and, as a corollary, update your virus definitions frequently.

o

To access your shared folder, you have a couple of options. You can either memorize your IP address, typing “\\” followed by the IP into the address bar in Internet Explorer, or you can map it as a network drive. To map a network drive, open My Computer, click Tools, and select Map Network Drive .

it as a network drive. To map a network drive, open My Computer , click Tools

Page 7

it as a network drive. To map a network drive, open My Computer , click Tools

Specify an unused drive letter and enter the network path for the share (ie. In the format of \\“Computer Name”\“ Folder Name” or \\“IP Address”). Once it has a drive letter, you access it just like a disk on the same computer.

o Check Reconnect at logon, so that the mapping will happen automatically every time you start your computer. Otherwise, you'll have to map it manually every time. Click Finish. The mapped drive is connected and appears in a new window. It should also be available in My Computer as a drive.

One Method for Protecting Your Shared Files Hiding Them Most likely, the “safety” afforded bySimple File Sharing is not enough for you. It’s not incredibly safe for both your computer and your files, as people would be able to add files to your shared folder willy-nilly.

One answer is to create a hidden share by adding a dollar sign ('$') to the end of the share name. A hidden share doesn't appear in My Network Places or Network Neighborhood on any of the networked computers. Only people who know the share name should be able to access it.

o

To create a hidden share, right click the disk or folder and select Sharing and Security. Specify a share name that ends with a dollar sign. Once again, use 1-12 characters (1-11 before the dollar sign).

o

Here we labeled the share Design$, but if you think anyone you don’t want to give access is clever enough to guess a name like Design$, use a more secure name, like a combination of letters and numbers.

o

To access your folder, either map the hidden folder as a network drive like before, or type in your IP address, followed by the name of the share. (For example,

\\192.168.0.4\Design$).

name of the share. (For example, \\192.168.0.4\Design$ ). One last point that we can never stress

One last point that we can never stress enough is to make sure that, if you share a folder at home, you constantly protect your computer from viruses and worms. If you already have an anti-virus program, such as Symantec Anti-Virus, this should be quite sufficient for protective means. However if you don’t have it, you can download it from the CIT website: http://www.cit.cornell.edu/services/nav/. You only need to enter your netid and your password. Make sure you update your virus definitions every two or three weeks.

Page 8

Chapter 2: Aspen Plus Simulation Engine Multi-Component Flash Drums

The purpose of this chapter will be to (re-)introduce you to the vile world of Aspen . It first goes step -by -step through the basics of Aspen, then specifies an important piece of equipment you should need to know how to model in Aspen, and covers how to perform a sensitivity analysis. The sensitivity analysis is instrumental in observing how process performance changes with varied operating conditions. Problem Statement:

A flash chamber operating 80 o C (176 o F) and 500kPa (72.5 psia) is separating 1000 kg/hr (2204 lb/hr) of a feed that is 10 mol% Ethane, 5% Propane, 15% n-Butane, 10% n-Pentane, 12% iso-Pentane, 8% n-Hexane, 30% Heptane and 10% Nonane. What are the product compositions and flow rates? Feed conditions are the same as that of the flash chamber. (Example taken from http://chemeng.stanford.edu/~charles/cheme120/Lectures/Lecture6-BubbleDew.ppt)

Solution Methodology:

The Basics (This section is a step-by-step process for those who have never used Aspen):

Starting Up:

First we shall discuss how to open Aspen.

o

Click on the .
Click on the
.

button, then click on

how to open Aspen. o Click on the . button, then click on and o This

and

o

This will bring up the Aspen startup screen. If this is the first simulation you have ever made, choose to create a Template, click OK, and

then select

choose to create a Template , click OK , and then select . The purpose of

. The purpose of this is to have all default settings in English Units. Click OK.

to have all default settings in English Units. Click OK . Setting up your flowsheet: First

Setting up your flowsheet:

First you need to select the piece(s) of equipment you desire.

o

To get to flash drums click on the Separators tab on the bottom of the screen. Then click the down arrow next to the picture labeled Flash2. Some choices of drums come up; select the top right by clicking on it. o Note: When you hold the cursor over each directory like “Flash 2” or Flash 3”, an explanation appears under it. All the selections in a given directory are exactly the same program and only LOOK different so that your flow diagram can look the way you want it.

o

Now click the mouse in the middle of the screen, and the drum should appear. Right clicking stops you from placing drums on the screen.

Page 9

o

To name the piece of equipment, left click on each piece(s) of equipment, right click and then select Rename Block. An easier alternative is to hold CTRL and type M. Type in an 8 character name for your unit, such as FLASHDRM for example.

o

Note: If you don’t want names to be shown on the flowsheet, click on the name, hold CTRL and type H.

o

Next you will need to connect streams, which is done by first clicking on the Material Streams tab at the bottom right of the screen. For

the feed, click and hold away from the feed side of the drum and pull the mouse to connect with the Red feed arrow of the drum (it’s on the left side).

o

For the exiting streams, click and hold the arrow on the top of the drum and pull the stream off the drum and release. Follow the same procedure for the red arrow on the bottom of the drum. Name streams using the same method as for naming the drum.

Note: Holding the mouse over the colored arrows, blue or red, will tell you the purpose of that stream (feed, tops, bottoms, etc) For your reference, our flowsheet is shown in the diagram above.

o

o

Click the

o Click the ( Next ) button in the middle of the top of the screen.

(Next) button in the middle of the top of the screen. Click OK to proceed to the next step

Setting Properties

o

To add a title, type it in the Title block that appears then click

.
.

o

Note: It’s often more useful to have results with components in the form of mole fractions instead of mole flow rates. To do this, under the Setup folder, simply go to Report Options, then the Stream tab, and choose Mole under Fraction Basis.

o

Now you need to add all of your components. Simply type in the name if you know it, or you can search for it using the Find button and typing in either a component formula or name. Unfortunately, you might have a little trouble finding iso-Pentane, since Aspen has its own

conventions for carbon chains with branching. To locate it, use Find, and type in butane, or search using its formula. The component with a name of 2(or 3)-Methyl-Butane fits the description of iso-Pentane, so press the Add button to add it to the list of components. Continue adding all components in our column and press Close.

o

Note: You can rename the components by changing their Component ID, and choosing Rename at the screen that pops up. When you

 

are done, click

.
.
 

Choosing an Appropriate Thermodynamic Property:

 

o

Now, you need to choose an appropriate thermodynamic property for the specific separation. Doing this can often be confusing, but don’t stress over it. Located in Appendix A is the very handy Thermodynamic Decision Tree. It gives you numerous options for deciding on a calculation method. Just start from the first branching (polar vs. nonpolar) and continue along it.

o

Note: For most problems that you face in your college career, you will rarely have to worry about your system having an electrolyte or

not having available interaction parameters. It is probably best to just assume thatAspen has the interaction parameters, unless you are dealingwith a very complex system or one that has little research associated with it. Our system has a number of nonpolar components, so start from there, and continue on.

o

Note: Here are some tips from Aspen on choosing a Thermodynamic system when you really don’t know what to use:

 

1. If you have a mixture of liquids that separate into two or three layers or if the liquids are non-ideal use an activity-based system. NRTL- RK, according to Aspen, is the one to pick if you have no real data to check your selection with.

2. If there is a dissolved gas in the liquid phase (as from sparging a gas into your reactor, etc) designate that gas as a Henry’s Component.

3. For all other systems, as a selection of last resort, use RK-SOAVE. This is an equation of state model.

 

For this sp ecific operation, I chose to use PENG-ROB, because it fits our “thermodynamic path.” PENG-ROB is usually a good choice if you have a system with a several hydrocarbons. With water in your system, things can get complicated. If water will separate out as a third

immiscible phase, don’t worry about it; but, if water is miscible (ie. water and ethanol), you will need to use a polar calculation method, such as NRTL- RK.

o

Important Note: The only real way to tell whether you have selected the proper system is to compare a T-x-y or x-y diagram generated by Aspen with REAL DATA. We cannot stress this point enough. If you have a relatively simple system, or one that has been researched thoroughly, you can probably get real data from numerous sources. This includes one of your professors, the infamous Perry’s Chemical Engineer’s Handbook, other library resources, etc.

Click

Click when you are done.

when you are done.

 

o

Enter the stream data and conditions for the feed stream (Temperature, pressure, flowrate, and composition). For having the composition in

o

terms of mole fractions, you must select Mole-Frac from the menu on the right. Click

Set the Temperature and Pressure of your Flash Drum, and again click on machine do the work.

of your Flash Drum, and again click on machine do the work. . . Click OK
of your Flash Drum, and again click on machine do the work. . . Click OK

. . Click OK to run the simulation, sit back and let the

o

To view your results click on the

button towards the upper right of the Control Panel that should now be up.

button towards the upper right of the Control Panel that should now be up.

o

Click on

Click on

and then you can scroll down to see the results.

 

o

Note: Any time you change conditions and want to run again you must Reinitialize the process before running. This is done by clicking

 

the button that looks like

.
.

o

In order to print your results, you can print your data table summary onto the flowsheet by clicking on the Stream Table button.

o

Assuming you have input the data correctly, you should have results similar to the data shown on the next page:

Page 10

These results are pretty close to what you would obtain using a Depriester Chart, so

These results are pretty close to what you would obtain using a Depriester Chart, so let’s move on.

Performing a sensitivity analysis:

If you wanted to determine how a component (say Ethane) in one of your output streams (let’s use the Vapor stream) is affected by the temperature, you could perform a sensitivity analysis showing how the flash changes while varying temperature.

Methodology:

o

Click on the Data Browser button

:
:

o

Open the Model Analysis Tools “folder”. Then click on Sensitivity.

o

Create a new sensitivity object by clicking on the

o Create a new sensitivity object by clicking on the button.

button.

Name it something you will recognize or just keep the defaultS -1 name given to you.

o

Then, under the define tab, create a new “Flowsheet Variable” by clicking on the New button. Create a name and specify what kind of variable it will be. Here we named it ETHVAP. o Note: Think of the flowsheet variable as that which is watched as Aspen performs its calculations, so it would likely be either mole fraction of Ethane in the liquid or vapor streams.

o

This might be a little confusing at first, so, as an example, look at our example in the diagram to the right:

o

Under the Vary tab, you are now asked to input conditions for the “Manipulated Variable,” which is the item Aspen changes while it performs its iterative calculations. This should be what you want to increment so that you obtain a change in the flowsheet variable, which will be the block variables temperature or pressure.

o

Choose a range of values over which the temperature will be changing. If you are performing an simulation where you want to show all the results in a range, merely choosing a range both above and below the initial guess, then narrowing the range down can be quite useful. I recommend using 100 to 300 as the temperature range to look at, and incrementing by 1. Again, if you are a little unsure of what this screen should look like, to the right is an example.

Page 11

by 1. Again, if you are a little unsure of what this screen should look like,
by 1. Again, if you are a little unsure of what this screen should look like,

o

Under the Tabulate tab, type in a column number, and enter the name of the Flowsheet Variable that you wish to evaluate.

o

Click the Next button to run the sensitivity simulation.

o

To view the results click on the Check Results button

o To view the results click on the Check Results button on the control panel, once

on the control panel, once the calculations have been completed.

o

Click on the Model Analysis Tools folder, then Sensitivity, and S -1 (or whatever you named the object) and then you can scroll down to see the results. If you want to graph the results, it is recommended that you highlight the data, copy (CTRL-C) the data then paste (CTRL-V) it into Excel or another database program. You should be able to graph it quite easily from that point on. o Note: To highlight a whole column of data, merely left-click on the column’s label box. This should select all data below it. If you want to highlight all columns in one click, just click the blank box in the top left corner of the Summary table. This is usually located next to the Status column.

As always make sure your answer makes sense. If your distillation column consists of 7 trays, and operates between -50 o F at the top and 700 o F at the bottom, you probably should take a closer look at your column.

This ends the first chapter on using Aspen . From here on, the manual will no longer be written in a step -by -step -by -step fashion, mostly because we feel you do not need to be led by the hand through the entire process. If there are potentially confusing instances, we will try to accommodate by including a visual aide. However, the main purpose for this manual is more geared to hopefully providing useful hints on managing the rather confusing simulation tool we refer to as Aspen.

Page 12

Chapter 3: Aspen Plus Simulation Engine Generating Graphs in Aspen

Now might be a good time to demonstrate how to use Aspen to create and edit pure component, binary and tertiary system property graphs. It’s a pretty useful tool to use if you need to give your manager hard-to-find physical property data for your system, or if you want to evaluate your system’s temperature and composition data before you perform a simulation.

Pure Component Systems:

Problem Statement:

Assume you are running an experiment using a simple pure substance (>99.9%), such as methanol (g) at 68 o F, and you require a few bits of information. You have found out through your various reference sources or your employer that, at atmospheric pressure, methanol boils at 65 o C (149 o F), so you will be required to heat your stream of methanol (obviously without exposure to air).

However, your experiment requires you to gather some information, for purposes of data verification, about methanol over the change in temperature from 68 o F to 165 o F. This includes documenting the constant volume heat capacity, the change in Gibb’s Free Energy and Enthalpy, and density over the specified range; as such, you will need to use Aspen to graph these physical properties of methanol.

Solution Methodology:

First you need to input the components and specify which thermodynamic system you are going to use.

o

Much like in the previous chapter, open either a new template or blank document in Aspen, and then add a piece of equipment, including the streams.

o

Open the data browser, click on the Components folder, and then go to the Specifications tab. Add all component(s) to the system. In our case, this is merely methanol.

o

Choose the thermodynamic system you will be working with by using the Thermodynamic Decision Tree, located in Appendix A. To input and edit your choice, click on the Properties folder. Then, click on the Specifications tab, and choose the system from those listed under Base Method. Here I chose UNIQUAC, but you could also choose NRTL, or WILSON. UNIQUAC usually provides a reasonable amount of accuracy, but the only way to be completely sure of a correct choice is comparison to real data.

Now we need to actually make the graphs.

o

Click on the Tools button located on the toolbar at the top of the screen. Click on Analysis and choose the option of Property. Then choose Pure.

o

When the Pure Component Properties Analysis Screen pops up, you need to choose the property you wish to analyze. From the problem statement, you are asked to plot constant volume heat capacity, t he change in Gibb’s Free Energy and Enthalpy, and density over the given range. The screen might look something like this:

in Gibb’s Free Energy and Enthalpy, and density over the given range. The screen might look

Page 13

Hit the Go button to plot the results of your analysis . With the plot done, you now have a couple of options.

o If you are satisfied with your plot, you can perform another plot analysis using a different property, or you could edit the way the graph looks by playing around with the Plot Wizard.

the graph looks by playing around with the Plot Wizard . Optional: If you want to

Optional: If you want to edit the plot:

o

Close the plot, and click on the Plot Wizard button on the analysis results screen. Click Next, then choose the plot type (there is probably only 1 or 2 options) and go to the next screen.

o

Select the data you wish to plot and click the > button to add data from the left column, < to remove data from the right column.

o

On this screen you are also given the option to change the units that are displayed.

o

On the next screen you can alter the title of the graph, and you can alter the title of the axes (Usually you would like to have the axis display the units, so do not remove %Unit%).

o

You can also change the type of gridlines that are present on the graph by changing the Grid Type or Line Type selections. When you are finished, click the Finish button. Our plot is shown in the diagram below.

o

With the editing done, you can print the new plot, and repeat the process to create plots for the remaining properties asked for.

D
D

o Note: With each type of Aspen graph, you will be presented more editing options when you right-click on the diagram. You can Zoom in or Zoom out, change the color scheme, etc. However, since these options are mostly for aesthetic purposes, I haven’t covered them.

Binary Systems:

Problem Statement:

Now assume that you are running a similar experiment, however you now have a mixture of Methanol, and n-Butane (50-50 mixture), and you wish to obtain an xy, Txy, and KVL diagram of your system.

Solution Methodology:

o

First you need to add n-Butane to your components list, and make any necessary changes to your thermodynamic system, if any. Remember this is done by going to the “Properties” folder, then hitting “Specifications”, and changing your Base Method. o Note: Whenever you make a modification to the current system, make sure you make any pertinent changes, especially if you are adding a complex component, or if your system is no longer at a low pressure, etc. Some other possible modifications could include adding Henry’s components, electrolytes, petroleum free-water method calculations, etc if you were dealing with the appropriate binary mixtures.

o

Click on Tools, then click on Analysis and Property. Under “Property” you should have the choice of Pure and Binary. Here instead choose “Binary.”

Page 14

o

o

o

o

o

o

o

When the Binary Analysis screen pops up, go the Analysis Type option, and you should have the choices of Txy, Pxy, and Gibb’s Free Energy of Mixing. Choose Txy.

Make sure you are operating at Atmospheric Pressure, and that your property options (Base Method, Henry’s Components, etc), range and number of points/increments are how you like it.

When you are done click Go.

The plot should pop up, and it should look like the diagram to the right:

pop up, and it should look like the diagram to the right: You can either edit

You can either edit the plot or go on to the next chart. In any case, go to the Binary Analysis Results window created along with the plot, and choose Plot Wizard.

If you want to edit the current plot, you would select the Txy type of plot. You would now have the same options for editing a plot as you would in a pure component situation. After making your diagram look the way you want it to, plot the new graph.

The other option is to choose either the xy or KVL type of diagram. When you choose either of these, you enter the plot -editing pages, which have already been discussed. When you are done editing, you can plot each graph.

Tertiary Systems:

Problem Statement:

Now assume that you are running a similar experiment, however, this time, some of the methanol in your system is actually ethanol (30-20 and 50 butane mixture), and you need to generate a ternary diagram of the system.

Solution Methodology:

o

Again, add the new component to your system and make the necessary changes to your thermodynamic system, if any.

o

Click on Tools, then on Analysis and then Property. Under Property you should have the choice of Pure, Binary and Residue with the addition of a third component. Here instead choose Residue.

o

When the Residue Analysis screen pops up, check to make sure that all of the options chosen are correct. Make changes as needed and select the number of curves you wish to generate. When you are done click Go.

o

The plot should pop up, and, depending on the number of curves you chose, it should look something like this:

of curves you chose, it should look something like this: I think the diagram looks fine

I think the diagram looks fine the way it is. There isn’t much in the way of editing options, other than the ability to add a legend, and modify the axis labels, so your graphing task should be complete.

Page 15

Chapter 4: Aspen Plus Simulation Engine Modeling Distillation Columns

Distillation columns will be the major units of your overall process flowsheet. Setting up a column is very similar to that of flash drums, but optimizing becomes much more complex. To guide you through some of the more useful simulation techniques, we’ll walk you through a quick example.

Problem Statement:

It’s your first assignment at your new job at the AMC ® Refinery, and you’ve been assigned to design a series of columns to remove propane. Propane tends to build up in reactor units downstream, and it is also saleable. For these reasons, you will need to first design a column to remove ethane, so that the saleable propane only contains 1% ethane. Next, your depropanizer column needs to remove 99% of the propane. You are given the following information about your stream, which enters your process at 100 o F and 250 psia:

Solution Methodology:

Component

Ethane

Propane

i-Butane

n-Butane

i-Pentane

Total:

lb/hr

lbmol/hr

1,804

60

12,214

277

16,100

277

28,887

497

938

59,943

13

1,124

Mole %

5%

25%

25%

44%

1%

100

The following instructions will guide you through the early steps of designing your deethanizer.

o

Add a column by selecting the Column tab at the bottom of the screen. RadFrac is the most common method we use to model our

columns, so click on the arrow next to RadFrac and select your favorite picture of a distillation column. Your selection really doesn’t affect the way Aspen does its calculations.

o

However, note: If your column does not have a reboiler or condenser, pick the picture that most closely represents your column.

o

Place the column on your flow sheet and name it. Now add your material streams and name them. Feed, Tops and Bottoms or some variant of them usually works. Remember if you are ever unsure of what a stream on the block (your column) represents, hold the mouse over the red or blue arrow and Aspen will tell you.

mouse over the red or blue arrow and Aspen will tell you. o Click Next and

o Click Next and then click OK to go to the Data Browser.

o Another hint: By now, you should realize that whenever tabs display

hint: By now, you should realize that whenever tabs display , it means you need to

, it means you need to enter more information before

display , it means you need to enter more information before , you can press moving

, you can press

moving onto the next step. Once those nasty red symbols turn into encouraging blue checks

the next step. If you have entered all information and need to go back to something else, feel free to pick the section you desire on

the left bar of the screen. You can also navigate backwards and forwards by clicking on the appropriate arrow:

to proceed tothe left bar of the screen. You can also navigate backwards and forwards by clicking on

.
.

Page 16

o

Supply a title if you wish.

Add your two components, remembering that if you

Next, in the Streams Data Browser, enter the

In the Blocks Data Browser, choose a number of

remembering that if you Next, in the Streams Data Browser , enter the In the Blocks

o

can’t type in the name of the component, you can search for it using the find options. As always, the next step is to choose a thermodynamic system from the Thermodynamic System Tree . We are going to use PENG-ROB for our example, and choose it under Property method, as shown at right.

o

conditions and flow rates of your feed stream (Remember: Make sure that all of the components add up or else you will get horrendous errors later on! We suggest using mole fraction because it is easier to make sure that the total is what it should be, since it should always be 1.)

o

stages, condenser, distillate rate , and reflux ratio.

o

Note: The condenser counts as the first stage, and the reboiler counts as the last stage.

o

For a good starting point on column properties, use the FACT Method. The FACT Method will estimate the number of trays and reflux ratio, based on the relative volatilities of components you want to separate, providing an excellent starting point. To refresh your memory, the number of actual distillation trays is:

 

<95%

< 98%

< 99%

99.5%

Product Purity # of actual trays

17/(a-1)

19/(a-1)

22/(a-1)

25/(a-1)

 

And the reflux ratio is:

 
 

Value of a-1

>1

+/- 0.5

+/- 0.15

R/D

1

2.5

6

o

Using FACT to start from, the relative volatility, a, estimated by the vapor pressure ratio of ethane and propane, is approximately 4. This indicates that our reflux ratio should be around 1. Next, we recommend usingFigure 19-19 in the GPSA Handbook to get a more precise

starting set of conditions, which, for a deethanizer, is 30 stages, a reflux ratio of 1.5. We also chose a distillate rate of 60 lbmol/h, since we are rejecting approximately that amount of ethane.

o

Now specify where the feed(s) enter the column. If you are feeding to the bottom of a column (below the lowest tray) you must change the Stage Convention from Above -Stage (the default) to On-Stage. We’ll choose to feed at tray 15.

o

Specify the column and/or condenser pressure. Setting only the condenser pressure is fine.

 

o

Note: Make sure the units you are using are correct. Most of the time you will be dealing with absolute pressure, so if you tell the column that it is operating at 10 inches of water it will treat the system as a vacuum. In other words just make sure you are using absolute pressure (unless the units on the right say psig or barg).

o

The condenser pressure dictates the column pressure, as well as the temperature of the condenser. You typically want to choose a condenser pressure to be the vapor pressure of the component you want to remove at around

You typically want to choose a condenser pressure to be the vapor pressure of the component

20

o greater than your cooling temperature.

o

Looking at Figure 23-20 in the GPSA Handbook, the vapor pressure of ethane at 120 o F is over 1000 psia! If we were to set the condenser pressure that high, too much of the heavier components would vaporize. The separation we are making is mostly between ethane and propane. The vapor pressure of propane at 120 o F is around 250 psia. Therefore we want to pick a pressure that is reasonably in- between these two pressures.

We’ll try running at 425 psia, and now the column should be ready to Run. Our results are pictured to the right:

o

Our separation converges, always a good starting point. We are removing 35% of ethane, but only losing 3% of

Page 17

propane out the tops. But let’s see if a couple changes can improve this separation to our target point.

o We can try changing several things to improve our column: reflux ratio, distillate rate, number of trays, and feed tray.

Optimizing reflux ratio:

o

o

o

o

A good place to start our analysis is with the reflux ratio. We want to find the smallest rate that satisfies our design spec of 99% recovery of propane in the tops. We’ll do this by performing a Sensitivity Analysis.

Remember that this is done by first selecting the Model Analysis Tools folder, then selecting the Sensitivity Analysis folder and pressing New to define a new variable. It’s fine to call this S -1.

Next, define the flowsheet variable of interest; again, this is the variable which you are observing while another variable is varied. We’ll define two: ETH as the molar flow rate of ethane in the tops, and PROP, the amount of propane in the tops (just to keep track).

A Variable Definition window should pop up that looks like this:

Definition window should pop up that looks like this: o o o o o Again, if

o

o

o

o

o

Again, if the variable of interest is a block variable (temp, pressure, heat duties) select Blocks on the left and then choose the Type, Block, and Variable. For Stream, blocks follow the same procedure but selectStreams on the left and then select the Type, Stream, and Component.

Similarly, in the Vary tab, choose your manipulated variable. Select what you wish to vary by selecting the variable from the pull-down menus on the left (molar reflux ratio is MOLE-RR).

Now either create a list of values for your variable or pick a range and a number of points or increments. We’ll vary the reflux ratio from 1 to 10, incrementing by 0.25.

If you wish to have more than one variable select the Variable Numberpull-down menu and select New. Repeat the process above until you have all of your variables.

o Note: The more variables or points you have significantly increases the amount of time it takes to run the simulation so choose wisely.

In the Tabulate menu, for each column write a number and then type in the name of your variable.

o Reminder: Instead of typing in the name, you can select the variable from a list. First right click in the variable space and then select Variable List. Double click on the variable you want to tabulate then close the Defined Variable List window. Right click again in the variable space and paste your variable name.

then click

Make sure you have reinitialized by clicking

% Component Recovery vs Reflux Ratio

100 90 80 70 60 50 40 Ethane 30 Propane 20 10 0 0 2
100
90
80
70
60
50
40
Ethane
30
Propane
20
10
0
0
2
4
6
8
10
12
% Recovery

Reflux Ratio

10 0 0 2 4 6 8 10 12 % Recovery Reflux Ratio o o o

o

10 0 0 2 4 6 8 10 12 % Recovery Reflux Ratio o o o

o

o

to run your simulation.

To view the results of the sensitivity analysis, in the results browser menu select Model Analysis Tools, Sensitivity, S -1

(if this is what you named it), and finally

A table of your results should now appear. You can either scroll down the list in Aspen or select and copy the desired columns and view them in Excel. The latter is useful if you wish to plot your results for visual comparison.

.
.

Shown to the right is a plot of the % recovery of Ethane and Propane in the tops vs. the Reflux Ratio:

Page 18

o

It looks like we might have a tough time recovering a lot of ethane with a low reflux ratio.

o

Note: we have to reject at least 80% of the ethane to meet our 1% ethane in propane spec. The graph indicates that above 8 there isn’t much change in ethane recovery when increasing the reflux ratio. We’ll set it at 8 for now and see if after we change some other specs we can lower it a bit.

% Component Recovery In The Tops vs Distillate Rate

100 90 80 70 60 Ethane 50 Propane 40 30 20 10 0 58 60
100
90
80
70
60
Ethane
50
Propane
40
30
20
10
0
58
60
62
64
66
68
70
72
74
76
% Recovery

Distillate Rate (lbmol/hr)

Optimizing distillate rate:

o

Increasing the distillate rate is not helping us very much – for every increase, we are letting at least as much propane into the

tops as ethane.

o

We should keep the distillate rate at just over 60, and see what else we can change to improve our process. With the distillate rate at 61 lbmol/hr, 96% of ethane is rejected out the tops.

o

Lastly, we’ll try to find the best number of trays and feed tray.

This is done by running a design spec on the feed tray for

several different sized columns. In doing so, we found that increasing to 35 total trays and feeding on tray 10 gave the best separation (results below), and also allowed us to decrease our reflux ratio to 5.75:

o

The ethane rejected is around 98%, and we are only losing 1% of the propane – Not bad! Since our overall goal is propane recovery downstream, and the ethane in the bottoms is around 0.4% of the propane in the bottoms, we should be safe with these numbers, but…

o

Note also that the tops temperature is 52 o F, which will require very expensive refrigeration. To increase the tops temperature, we can adjust the column pressure, reflux ratio, and distillate rate.

o

First, we ran a design spec on the reflux ratio to obtain an ethane molar flow rate in the bottoms of 2.7 lbmol/hr (approximately 1% of propane). This gave us a reflux ratio of 2.2.

o

Though we rejected our desired amount of ethane, the tops temperature was still too low. To try and increase the temperature, we increased the condenser pressure to 460 psi. This caused more ethane to leave through the bottoms.

o

Lastly, we incrementally increased the distillate rate until the amount of ethane in the bottoms was nominal. The more total moles you let out the tops, the more ethane will be rejected through the tops. The tops temperature ended up at 113.5 o F:

the tops. The tops temperature ended up at 113.5 o F: o Originally, we used the
the tops. The tops temperature ended up at 113.5 o F: o Originally, we used the

o

Originally, we used the FACT Method to calculate our reflux ratio to be around 1, but we ended up with a reflux ratio of 2.2. The vapor pressures of ethane and propane are so different that a ends up being huge, which means that it should be an easy separation with a low reflux ratio. Instead, we’re seeing a hard separation, and needing to use a higher reflux ratio. This is probably because there is so little ethane in the feed that the separation is actually more like a stripping column than a distillation column.

o

The message from this is that FACT is a good place to start from, but you will still need to be aware of the specific circumstances of what you are trying to model.

Page 19

o

Next, we will design a depropanizer to remove the saleable propane from the deethanizer’s bottoms stream. The composition of the bottoms stream is:

 

Component

lb/hr

lbmol/hr

Mole %

Ethane

34

1.1

0.1

Propane

12,078

273.9

25.7

i-Butane

16,093

277

26

n-Butane

28,887

497

46.8

i-Pentane

938

13

1.4

Total:

58,030

1,062

100

o

We will design the column in the same way as we did before, in a separate file, and later we will join them together.

o

We will start with 35 trays, and feed on tray 15. We’ll set the reflux ratio to 3, the distillate rate to 275, and the condenser pressure to 200 psia, which is the approximate vapor pressure of propane at 120 o F. The results from this initial simulation are:

120 o F. The results from this initial simulation are: o These results are very good
120 o F. The results from this initial simulation are: o These results are very good

o These results are very good for an initial simulation. We’re recovering 94.5% of the propane, and the tops are 94% pure. Let’s see if we can improve the purity of the tops to less than 1% impure.

Since the overheads of this column will be a lot greater than the overheads of the deethanizer, there shouldn’t be any problems with a low reflux ratio. Let’s try to find R D .

Finding R D :

In the beginning stages of process design, you are not expected to know anything definitive about number of trays, reflux ratio, heat duties, etc. You can tentatively set R D to R MIN + 10%. We’ll try playing with some Design Specs to find R D .

o

The minimum reflux ratio can be estimated by increasing the number of trays dramatically, and changing the feed stage to get the best separation. Note the reflux ratio required for the best separation, and repeat until increasing trays does not change the reflux ratio. This will be your R MIN .

o

To perform this analysis, go to Design Spec under Flowsheeting Options in the Data Browser.

o

We’ll define two specs, setting the mass flow rate of propane in the tops and bottoms to our desired separation.

o

We’ll then vary the reflux ratio (our example will choose 1 to 10) and the distillate rate (we’ll try 274 to 350 lbmol/hr), and increase the number of trays until the reflux ratio satisfying our specs does not get any smaller. Our best results are shown in the Results Summary to the right:

o

With 600 trays, feeding on stage 100, our reflux ratio is 3.

R D is then approximately 3.3.

o

Plugging in 3.3 as the reflux ratio, our stream results are (right):

o

These results are excellent – 98.9% recovery of propane, and 97.8% purity. We can try improving these results by performing a Sensitivity Analysis on distillate rate, feed tray, and total number of trays.

o

After checking out feed tray and number of trays, we get the best results with 35 trays, feeding on tray 11.

o

Lastly, we will try doing a Design Spec, to see if we could improve our results. Our two specs will be 99% recovery of propane in the tops, and 99% molar purity of the tops. We’ll vary our distillate rate and reflux ratio from 273 to 350, and 1 to 10, respectively. Here are the results:

Page 20

With a distillate rate of 274 lbmol/hr and a reflux ratio of 4.2, we have
With a distillate rate of 274 lbmol/hr and a reflux ratio of 4.2, we have

With a distillate rate of 274 lbmol/hr and a reflux ratio of 4.2, we have obtained our design specs!

Connecting Columns:

o Now that we’ve optimized our two columns, where the feed of the 2 nd is the product of the 1 st , it is very easy to connect them. All you have to do is select the picture of one column, copy it, and paste it onto the other flowsheet:

one column, copy it, and paste it onto the other flowsheet: o Next, right-click on one
o Next, right-click on one stream, we’ll Reconnect Destination of BOTDEETH to DEPROP. Arrows should
o
Next, right-click on one stream, we’ll Reconnect Destination of BOTDEETH to DEPROP. Arrows should show up on various locations
where you can reconnect the stream to. We will reconnect it to the feed of DEPROP, the depropanizer. We can then delete PROPFEED:
o
You may then reinitialize and run the simulation; Aspen may ask you for a new feed stage for the second column. Enter the same stage you
had been operating at previously. In addition, you might need to add a heat exchanger or pump in case what you took from the first column
is too hot or cold for what you designed for the feed of the second column or the pressure of the second feed doesn’t match the pressure of
the first bottoms product. Here are the results of the combined columns:

Page 21

o The two columns now produce saleable propane with 99% purity . We can finally

o The two columns now produce saleable propane with 99% purity.

We can finally call this chapter to a fond end. We realize this is your first time attempt to model distillation columns. It's quite a lot of information to take in, but once you have read through and perhaps tried running this example yourself, you can officially call yourself an Aspen Guru.

Page 22

Chapter 5: Aspen Plus Simulation Engine Modeling Reactors

Eventually, all components of your design process will need to be modeled in Aspen so that a cost estimate in Icarus can be evaluated; this includes your reactor, too. In many cases you will not be able to model your reactor at all, but it still needs to be represented in Aspen. The simplest way of representing a reactor is by using RStoic type reactor. It is a reactor that separates components and performs reactions based on chemical stoichiometry.

The main purpose for this chapter, therefore, will be to give you an introduction to the “ black boxes” of Aspen. These are for stating material “x” goes into your device, and “x” (or “y,” in the case of reactors) comes out, and for having your equipment (simple reactors, mixers, splitters, and various solid separators) represented in Aspen. Each can usually be adequately modeled in Aspen, but for your design needs, you don’t need to worry about it, nor should you immediately have enough information to perform such a task. For right now, let’s work with a reactor.

RStoic Type Reactor You would usually use RStoic if the kinetics of the reaction are unknown. The only real requirements are that you know the stoichiometry of the reaction and the reaction’s percent molar conversion. RStoic can model reactions occurring simultaneously or sequentially and can determine selectivity and perform heat of reaction calculations. However, it is always recommended that hand-calculations of these are done before modeling, since there is no real basis for comparison.

Let’s start with an example. Let’s assume you were in one of the Tuesday ethanol plant design groups and you needed to model your reactor. The typical ethanol reactor operates according to the reaction:

C H

6

12

O

6

Æ

2(CH CH OH)

3

2

+

2(CO )

2

Assuming that you found in several journal papers or patents with verified documentation that the typical molar extent of reaction for an ethanol plant was 99.5% you should be able to model this reaction in Aspen. (I am making this number up for this example. If you are in a Tuesday group, you will need to find this number or more pertinent information yourself)

Solution Methodology:

o Open Aspen, select the Reactors tab and place an RStoic reactor on the flowsheet. Connect all required streams (there should be 2), and open the data browser.

RStoic reactor on the flowsheet. Connect all required streams (there should be 2), and open the

Page 23

o

o

o

o

Add a title if you like, change the report options (Setup Report Options Stream Tab) so that you have both molar flow and mass flow reported, and then add your components (there are 3: Ethanol, CO 2 and “Dextrose”). If you were dealing with a whole system, there would obviously be more components but these are the only ones pertinent to our little example.

Select your base thermodynamic method. I chose UNIQUAC, but it isn’t really too important for this example.

Specify the conditions of the feed stream. In this case we can assume that the feed is at normal conditions (68 o F, 1 atm), and that the flow rate is 150 lbmol/hr. Under the composition, there is only one component in the feed (Dextrose), so enter it as a 100%.

Next, you should be in the block options. Specify the temperature and pressure at which the reactor should be operating. You may again find, in a paper or journal, a typical temperature that the microorganism facilitating in the reaction operates best around is usually 95-98 o F; In that case, let’s use 98 o F as our reactor temperature. For the sake of having a number, let’s use 2 atm for the pressure.

o

o

o

You next need to specify the reaction. You create a reaction by clicking on the New button in the Reactions tab; if the reactor had multiple reactions occurring within, you create an item for each reaction. If the reactions occur in series you would need to check the box below the New button.

You need to now edit the stoichiometry of each reaction that is involved, and this means selecting each reactant or product, and supplying a reaction coefficient. Products have a positive coefficient, but reactants will be negative since they are being used up. (ie: dextrose = -1, ethanol = 2).

The final step is to specify the Molar Extent of Reaction or Fractional Conversion. Either works. The screen could look something like the diagram to the right:

o Run the simulation.

Optional : Finding the Heat of Reaction

the simulation. Optional : Finding the Heat of Reaction o You might need to specify the

o

You might need to specify the heat of reaction so that the heat duty can be accurately transferred to Aspen Pinch or Icarus.

o

The best way of doing this is either to have Aspen calculate the value (by going to the “Heat of Reaction” tab, selecting “Calculate heat of reaction” and input the reference component), or to perform the calculations by hand and enter the resulting value by selecting “Specify heat of reaction” and inputting it.

Here is a little rundown of the requirements for each type of reactor. For Design class you probably should just stick to either RStoic or REquil, and realize that no professor in Design class will seriously expect you to completely model all aspects of your reactor in ASPEN.

Reactor

Conditions/Restrictions

Output (Tab)

RStoic

o Reaction stoichiometry known;

o

Mass & energy balances (balance);

o

Reaction kinetics are not important or known;

o

Compositions of outlet stream (phase equilibrium);

o

Conversion is known.

o

Heats of reaction (reactions);

 

o

Product selectivities (selectivities).

RYield

o Unknown stoichiometry or kinetics;

o

Mass & energy balances (balance);

o

Yield distributions known.

o

Compositions of outlet stream (phase equilibrium);

 

o

Outlet component weight distribution factors (weight distribution);

o

Pseudocomponent results (pseudocomponent breakdown).

REquil

o Stoichiometry known;

o

Mass & energy balances (balance);

o

Some or all reactions reach equilibrium;

o

Equilibrium constants for reactions (K eq ).

o

OK for one- or two-phase reactors.

RCSTR

o Reaction kinetics known;

o

Mass & energy balances (balance).

o

Contents of the reactor have same properties as outlet stream (Assumes perfect mixing);

o

Specify temperature or heat duty, calculates the other.

Plug

o Reaction kinetics known;

o

Mass & energy balances (balance);

o

Cooling stream (cocurrent and countercurrent) optional;

o

Polymer distributions (distributions).

o

Rate-based kinetics only.

RBatch

o Reaction kinetics known;

o

Mass & energy balances (balance);

o

Rate-based kinetics only;

o

Polymer distributions (distributions).

o

Specify stop criteria.

Page 24

Chapter 6: Aspen Plus Simulation Engine Using RateFrac for Rate-Based Simulations

The purpose of this chapter is to get you acquainted with a new and powerful Aspen tool for systems such with rate-based models such as extractive and azeotropic distillation columns, absorbers and strippers, called RateFrac. We will run through a fairly thorough example, and we will try to give some useful hints for designing your absorber or stripper.

RateFrac is a rate-based nonequilibrium model for simulating many types of multistage vapor-liquid operations and simulates actual tray and packed columns, rather than the idealized representation of equilibrium stages. This is a relatively new item in your Aspen arsenal, so understanding how to fully use it can be complicated, but it shouldn’t be completely unfamiliar territory.

Problem Statement:

Let’s assume first that we have a packed tower absorber and acetone in a stream of air is being absorbed by water. The acetone composition in the inlet air entering at the bottom of the column is 3 mol%, and in the outlet gas leaving the top of the column it should be at most 0.5 mol%. The total entering gas flow rate is 15,000 mol/h, and the pure water inlet flow rate is 45,000 mol/h. You will need to find the outlet acetone composition in the water. The column specifications are as described below:

Temperature and Pressure: 68 o F, 14.7 psia Diameter: 1.5 ft Height: 7 ft

: 68 o F, 14.7 psia Diameter : 1.5 ft Height : 7 ft Solution Methodology:

Solution Methodology:

First we need to set up our example simulation.

o

First of all, open a blank simulation in Aspen. Under the Columns subdirectory, place a RATEFRAC column on the flowsheet. Pressing

the

the arrow next to the RateFrac block will give you a list of icons you can

arrow next to the RateFrac block will give you a list of icons you can use to represent the column. Choose any icon, but it is

recommended that you choose a picture that most accurately represents your desired piece of equipment.

o

Next attach all required Material Streams: this includes 2 inlet streams, one tops stream and one bottoms stream. If you have a more complicated system you could have to add side product, free water distillate, or pseudo streams. o Note: You are able to move the head of the stream arrow by clicking on it, and dragging it along the edge of the piece of equipment to another location.

o

Once the flow sheet connectivity is complete, click Next.

Page 25

o

Under the Setup folder, you can give your model a title, but more importantly you can change the output options for your results to a mole fraction basis.

o

Note: Remember this is done by going to the Setup folder Report Options folder Stream tab and de-checking mole in Flow Basis and checking it in Fraction Basis. Once you are done, click the Next button.

o

Enter your components. There should be 3 in this example: Air, Acetone, and Water. Usually you will need to Find your components in Aspen’s directory, but you can merely enter these names and the correct item should appear. Click Next when you are done.

o

Under the Base Method option, enter the Thermodynamic Method you will be using to evaluate the properties of your components. Again, if you are unsure at any time what your property method is, don’t hesitate to use the Thermodynamic Method Tree located in Appendix A. In our example, we will be using NRTL. Click Next. If the Binary Interaction screen appears, just ignore it, click Next again and choose to Go to Next required input step.

o

On the stream property sheets for each inlet streams, fill in the required fields according to the problem statement, including temperature,

pressure, flow rate, and composition. While inputting your values, remember that the molar flow rate is in metric units, and not the English Unit lbmol . Also, in order to enter your compositions in terms of mole fractions, you will need to change the composition basis from Mole- Flow to Mole-Frac.

o Note: In the inlet vapor stream sheet, if you were given vapor fraction and one other property, instead of both temperature and pressure, you can enter the percent vapor fraction under the missing property.

Now, we need to specify information for our Absorber. This includes specifying numbers of “segments,” defining packing types and sizes, and entering column dimensions.

o

Once you are done, click Next and the input sheet for your Block, the absorption column, should appear. Here you will need to enter the number of segments.

o Note: Aspen, for the purpose of easing calculations in evaluating heat and mass transfer rates between contact phases, divides columns into segments; in a packed bed column, these get translated into segments of packing, or a series of trays in a trayed column. More segments usually mean more accuracy, but usually you should stick to one segment per foot or two of packing. In our example, we will use 10 segments, but feel free to use more if it will be more accurate.

You will also need to select the Condenser and Reboiler type. However, since we are modeling an absorber, select the None option for both the condenser and reboiler, and click Next.

o Note: Because of this, the Operating Specifications at the bottom of the screen should have both condenser and reboiler duty set at 0, which is what we want.

Next, you will need to enter the pressure, which, as stated in the problem, is 14.7 psia for the first segment. The first segment is the first segment from the top, so first change the view to Top/Bottom if it isn’t already chosen, and enter our pressure as the Top Segment Pressure. When you are done, click Next.

You should be at the Tray Specs object manager screen. Since we are using a packed tower instead of a trayed tower, you can skip this

step by simply going to the

(ie. 1) and click OK.

Input the number of the last segment, which in our example is 10, under Ending Segment. Since the packing type was not specified in the problem, you are free to pick any type. We have chosen 1 in. Plastic Pall Rings, but if you wish to have a basis for comparison, there is a listing of information on various packing types in Figures 19-23 and 19-24 of the GPSA Handbook.

Next, we need to input an initial packing height that will be used to calculate our gas and liquid concentrations. This is given in the problem statement as being 7 ft, so when you are done, click Next.

On the next screen, enter the diameter given in the problem statement of 1.5 ft and click Next.

o Note: If we were given the percent flooding instead of the column diameter in the problem statement, we could have Aspen calculate the diameter for us by selecting Use calculated diameter. The only requirements are inputting the percent flooding, the number of the bottom segment, and an initial estimate for the diameter.

On the Material Streams screen, you need to specify where the segments the

folder in the data browser. Now, click the New button, and enter the starting segment number

o

o

o

o

o

o

button, and enter the starting segment number o o o o o o o feed and
button, and enter the starting segment number o o o o o o o feed and

o

feed and outlet streams are located. Since this is an absorber, the feed and outlet streams all leave either the top (Segment 1) of the column or the bottom (Segment 11) of it, with the gas entering from the bottom and the liquid from the top.

o Note: Notice that the number 11 was stated for the bottom segment. This is because the default convention for stream location is Above Segment. You can make it so that the bottom segment is 10 by changing the convention to On Segment as shown to the left.

You have entered all the information Aspen needs, so Run the simulation by clicking Next and hitting OK at the Required Input Complete screen.

You can access your results by clicking on the shown on the next page:

o

o

your results by clicking on the shown on the next page: o o button and choosing

button and choosing the Streams folder. They should look something like the picture

Page 26

Unfortunately, our resulting acetone composition in the exiting gas stream doesn’t meet the requirement set

Unfortunately, our resulting acetone composition in the exiting gas stream doesn’t meet the requirement set by the problem statement of 0.5 mol%. There are a couple of things you can do to correct this. Change a non-numeric estimate or a variable Create a Design Spec

Changing a Non-Numeric Estimate or a Variable The first option is probably the easiest to do if you have a result that is very close to what you want and you weren’t looking for an answer with a high degree of error tolerance. Though, as a warning, this does not always work to benefit you, so try it only when your results are very close to the spec. If this is not the case, do not attempt at all, since this will be very tedious and potentially unreliable.

In our case, this would be a good option, since we have at least one non-numeric variable, that being our type of packing. Any one of the options associated with it could be altered to theoretically give us a better answer; this includes size, type, and material. One possible solution is to change the Pall Rings from Plastic to Metal material. This results in an acetone composition of 0.4 mol% which satisfies the problem statement. But, for the sake of getting more accurate results, let’s try a Design Spec.

Creating a Design Spec

Creating a design spec for a RateFrac model is practically the same as creating one for a distillation column, so this should be a refresher course for those with distillation experience, or those who have already read Chapter 4. Our wish here is to vary the packing height in the column in order to obtain the desired mole fraction of acetone in the gas outlet stream.

o

Reopen the data browser and choose the Flowsheeting Options folder and then open the Design Specs folder.

o

Create a new design spec by clicking the New button on Object Manager Screen that pops up. Click OK to keep the name of DS -1, or change it if you are the discerning type.

o

Now we have to create a flowsheet variable, which will be the mole fraction of acetone leaving the column in the gas outlet stream. So, click on New, and name it something easily remembered, like ACETON and select OK. On the Variable Definition form, select the category All (or Streams), and fill out the required fields under Reference and click Next.

o

Under the Spec tab, enter the name of the variable just created, and enter under Target the value of the variable, which is 0.005. Finally, set the error Tolerance, which is the allowable difference between the spec and the target, and it is necessary to stop iteration. Here, we set the tolerance to 0.0001, though you don’t have to adhere to that number.

o

Under the Vary tab, you will need to specify t he manipulated variable, which, in our case, is the height of packing. Choose Block-Var, then name of your block ABSORBER, and the type of variable that Aspen is changing. In this case, since we are varying the total height of the packing, select HTPACK as the variable. o Note: Don’t get discouraged while looking through this giant list of strangely labeled manipulated variables. Remember that, if you are confused, you can highlight the variable and look at the dialogue box at the bottom of the browser to obtain a description of what each represents.

o

In the ID1 field enter the column number, which should be 1, and in ID2 enter the starting segment number, which should also be 1.

o

Next, choose a range for the upper and lower limits for your manipulated variable. A reasonable set of values for Aspen to iterate between is 30 (ft) as your upper limit and 1 (ft) as your lower limit.

o

Reinitialize and Run your simulation. You can view your results by clicking on Results Summary, then choosing the Streams folder to obtain the resulting stream compositions, or you can click on the Convergence folder to obtain your new packing height. This information should be listed under Variable Value as having a value of 8.06 ft of packing material if everything was done per our instructions. Notice that this is not the actual height of the absorber. You would need to account for the area above and below the equipment.

Page 27

Optional: Modeling Reactions Within Your Absorber Some design groups might have system that incorporates a stripping or absorbing column with one or more reactions occurring inside the model. If you have time and feel the need to have your process as detailed as possible, you might want to model the reactions occurring within your piece of equipment. Reactions are introduced to Aspen in a similar manner as they are in Reactors (Chapter 7).

o

After you have finished entering your packing material specifications, click on the Reactions sub-directory located below it. If you are unable to find this, look at the picture below.

o

Enter the starting segment and the ending segment, then, since we haven’t entered the reaction yet, create a new Reaction ID by choosing New from the pull down menu. Again, you can give the reaction a name, but its default should be R-1.

give the reaction a name, but its default should be R-1 . o Under the Reactions

o

Under the Reactions main directory, click on the newly created R-1 reaction. You will need to specify the stoichiometry of the reaction. So, simply click on New, and select the reaction type that fits your reaction. o Note: Usually you will be dealing with an equilibrium or conversion reaction, but a salt precipitation is not out of the question. In that specific case, select the precipitating salt from the list, then hit OK.

o

Now you need to fill out the stoichiometry form. Kinetic/Equilibrium/Conversion type reactions should yield virtually the exact same form as the one in the previous chapter, but if you are dealing with a Salt Precipitation, you merely need to specify the two reacting agents, and their stoichiometric coefficients, that precipitate out the salt.

o

With your Reaction sheet filled out, merely Run your simulation, and Aspen should model all reactions, as well as create your column.

This concludes our discussion on RateFrac columns. We hope that it has been helpful in modeling any absorber or stripper that you have in your design process, or, say, if you are furiously writing up your Stripping column report for UO Lab one or two days before it is due because you were unable to figure out how to get this part to run. In such situations, keep telling yourself “It’s not me, it’s ASPEN’s fault.” That always made my group feel a whole lot better.

Page 28

Chapter 7: Aspen Plus Simulation Engine Putting the Pieces Together

This will be our final chapter dealing solely with Aspen Plus, and as such, we have decided to give an example that tries to pull together most of the work and information gleaned from the previous chapters into one. We will try to show you how to model an entire process, which you could hopefully translate to your own work.

Remember that your system might be a little more rigorous than ours, or not have the exact same pieces of equipment, but it isn’t impossible to model it if you take what you have learned here and try to adapt it to your own specifications.

Problem Statement:

In the process pictured below, we have a stream of natural gas at 400 psig and 110 o F. This stream contains numerous saleable propane, butane, and C 5 + (pentanes and above), as well as a large percentage of methane, all of which could be sold at a higher price in its purified liquid state. To extract these, we first cool our feed to -20 o F by heat exchanging with our cold residue gas, and then with propane refrigerant, thereby causing condensation of the heavier hydrocarbons. The feed is then introduced to the bottom of an absorber, where chilled lean oil (nonane) is introduced at the top. The lean oil will absorb the heavier hydrocarbons from the gas. The methane-rich Residue Gas can then be resold at a higher price than purchased. The lean oil and hydrocarbons leave the bottom of the absorber, and are introduced to the top of a stripping column, where some propane and lighter hydrocarbons are removed by a vapor stream created by a reboiler. The stripper tops are recycled back and mixed with the feed, whereas the stripper bottoms are then fed into a depropanizer, where saleable propane is removed in the tops. The bottoms are then sent to a debutanizer, where saleable butane is removed. We could, later on in our process, remove any excess nonane from our process with another distillation column and recycle it back to the absorber, but let’s just stick with what we have now. Our goal in this example is to recover more than 85% of the propane in the feed. Its product stream can contain at most 1 mol% ethane.

feed. Its product stream can contain at most 1 mol% ethane. Solution Methodology: You may realize

Solution Methodology:

You may realize that you probably don’t have enough information to completely design this system. The absorption/stripping column can be pretty challenging, because, aside from not knowing the column dimensions, you also do not know the flowrates of the absorbing oil or the distillate rate in the stripping column. One of the best suggestions we can give is to use the GPSA Handbook, and find initial estimates for your specific column. Listed below are some good estimates for the first two columns.

Absorption and Stripping Column Specifications: (Partially suggested by the GPSA Handbook Fig. 19-20)

Absorber Height:

23 ft.

Stripper Height:

15 ft.

Diameter:

36 in.

Diameter:

18 in.

Packing:

2 in. Plastic Pall Rings

Packing:

1 ½ in Plastic Pall Rings

Setting Up Your Absorber:

Page 29

First, let’s setup our flowsheet. Our recommendation is to input one piece of equipment at a time, instead of putting in every piece all at once. It is easier to try to figure out why your column dries up with one column than with 4. Without going into details because this should be somewhat obvious, column optimization is also much easier separately. So, we are only initially going to deal with our first main column, the absorber, and not worry about the recycle or the stripper. However, since there is a recycle, it wouldn’t hurt to add a stream mixer before the feed, as shown below in our flowsheet. Trust me, it will save you time in the long run, when you have actually added the recycle stream.

long run, when you have actually added the recycle stream. Next, you need to setup the

Next, you need to setup the specifications for your column in the Data Browser (If you are confused on how to do this, or are unsure, see Chapter 7 for details on setting up your absorber column).

o

Enter all of the components into the component setup page, realizing that, if you are entering iso-Hexane you would need to Find using “Methyl -Pentane.” Make sure to also add n-Nonane.

o

Enter the Thermodynamic Property Method you will be using.

o

Note: As well as the recommended Thermodynamic Property Tree, an index of most common abbreviations for the models is given in the online help. However, the list is not complete. A complete list of abbreviations is given in the online manuals, accessible by going to Start -> Programs -> AspenTech AES 11.1 -> Aspen AES 11.1 Documentation. Look for the pdf file Physical Property Methods and Models. This manual will have details of the parameter names and t he models.

o

Enter your stream specifications. Remember that both nonane and feed streams will be at 400 psig and -20 o F. We don’t know what our flowrate of nonane will be, but we can take a fairly educated initial guess by using the suggested L/G ratio in the GPSA Handbook Fig. 19-19, and scaling it up for this process. A good starting guess for the mass flowrate of the nonane would be 1,600 lb/hr (this is a number we arbitrarily chose), but, as this number will be overridden later on, try not to marry yourself to any initial guess made.

o

Note: If you are unsure of the volumetric flowrate of your gas, you can find a pretty reliable online weight, flowrate and volume converter for crude oil, its products, and various hydrocarbons at http://www.processassociates.com/process/basics/oil_vw.htm, or one of many other online unit conversion websites. You could alternatively use Perry’s Chemical Handbook, but using a conversion calculator, such as the one on the website, will save you precious time.

o

Now you need to enter the number of segments and the conditions for your Reboiler and Condenser (you have neither, so choose None for both). More segments usually translate into more precision, but make your preliminary guess. We are using 10 as our number. o Important Note: From the problem statement, we are missing two important pieces of information: the column pressure, and the flowrate of our absorbing oil, Nonane. Since we are trying to optimize our system, while remaining within the specifications set by the problem, we will first need to choose one of these two variables to keep constant, while we optimize the column via the other variable. In our example, we chose to set the pressure to be our constant, and, later on, tried to find the “best” flowrate of nonane to obtain a desired

 

percentage of our component(s) retained.

o

So, first you will need to enter the pressure at the top segment, which can often be tricky. If you use a pressure higher than that of your feed streams, your segments could dry up because there isn’t enough driving force into the column, which will give you severe errors, or stop calculations entirely.

o Note: Since this wasn’t really stated in the problem, we are going to set some goals for the amount of some components in our packed columns. That said, we will arbitrarily try to achieve a minimum of 95% recovery of propane in the absorber, and between 85 – 90% in the stripper, so that when it gets to the DePropanizer, it has more than enough room to achieve the purity specification, and still maintain 85% recovery.

Page 30

o

Shown below is a diagram of how some important components will be retained in our absorbing oil when we vary the pressure of the column (We created this using a Sensitivity Analysis after completing our first run, so don’t worry about where this came from right now). o Note: I have only shown the effects on Methane, Ethane and Propane, since we are attempting to retain most of our propane, while getting rid of most of our lighter hydrocarbons. Our heavier hydrocarbons, you will find later on, should have an even better retention rate than propane (near 99%), so, for now, we will mainly be concerned with these three components.

% Component Retained vs. Column Pressure

100% 90% 80% 70% Methane 60% Ethane Propane 50% 40% 30% 20% 10% 0% 0
100%
90%
80%
70%
Methane
60%
Ethane
Propane
50%
40%
30%
20%
10%
0%
0
100
200
300
400
% Component Retained

Column Pressure (psig)

o

o

For now, let’s assume that the added benefits to our separation gained from using a lower pressure won’t change too dramatically upon changing the flowrate of our absorbing oil. You may notice that, with a column pressure greater than 80 psig, there is no appreciable loss in propane, however a significant amount of our lower weight hydrocarbons are removed. For our current flowrate of nonane (1,600 lb/hr), at 80 psig, we have removed 95.3% of the methane and 48.5% of the ethane, while still retaining 99.4% of our propane. This is a pretty good number to use.

Next, you need to enter your Packing S pecifications, Col umn Diameter, and Feed Locations (remember to use On Segment convention).

Now enter any specifications for the mixer. This includes the stream pressure and temperature conditions. After you are done, Run your simulation. Your results should look something like those shown below.

Your results should look something like those shown below. We have successfully removed a fair amount

We have successfully removed a fair amount of the methane and nitrogen, but some of the other components that we wanted to remove (ethane and carbon dioxide) are still present in relatively large amounts.

Page 31

We will minimize this by now running a sensitivity analysis on the flow rate of our absorbing oil, but you could take a possibly more difficult path by creating a design spec and solving for the flow rate of a specific component. The latter is potentially more complicated because you have the risk of going crazy from running your design spec and finding it doesn’t work, or it doesn’t get the separation you want, etc. You may take either path, however, we specifically chose to use a sensitivity analysis on the flowrate of nonane because we can see easily see how many components change with varying the nonane flowrate. (For more information on performing a Sensitivity Analysis, see Chapter 2)

o

Firstly, and always remember to do this before you run your simulation, Reinitialize. You will find out that if you don’t reinitialize you will get fatal errors, and Aspen will terminate its calculations. You probably won’t realize this at first, but if change your input and then run your fairly complicated simulation, Aspen will freak out! You will, at the very least, see warnings everywhere.

o

Next, go to your Model Analysis Tools folder and then Sensitivity Analysis. Under the Define tab, create a flowsheet variable and input the information for the first component you want to look at; here, you will want to find the “ Mole Flow” of Methane in TOSTRIP. When you are done, repeat this for the other two components we were looking at earlier (Ethane and Propane).

Under the Vary tab, you want to input the information for the Mass flowrate of Nonane. I would recommend varying the flowrate from 500-4,000, incrementing by 100. Finally under the Tabulate tab, enter the name of each variable and the order you want their spreadsheet column to appear as. Run the simulation. Our results are shown in the plot to below.

o

% Component Retained vs Nonane Flowrate

100% 90% 80% Methane 70% Ethane 60% Propane 50% 40% 30% 20% 10% 0% 500
100%
90%
80%
Methane
70%
Ethane
60%
Propane
50%
40%
30%
20%
10%
0%
500
1000
1500
2000
2500
3000
% Component Retained

Flowrate of Nonane (lb/hr)

o

By using a flowrate of Nonane of 1,000 lb/hr, we will retain 95.2% of our propane, 36.6% of the ethane, and 3.4% of the methane. Using this new flowrate, run the simulation once more as a check.

We have managed to optimize this column just by varying the pressure of the column and the flowrate of the absorber oil. If you want to try varying another aspect of this column (number of segments, packing type, temperature or pressure of the feed, etc) go for it. You might very well achieve a better separation. Using the new chosen flowrate, our results are as follows:

it. You might very well achieve a better separation. Using the new chosen flowrate, our results

Page 32

With your work on the absorber ostensibly completed, you will need to add your second piece of equipment, the stripping column. However,

since you need the feed going to both columns at 400 psi, you will require the addition of a pump as well. So, go back to your flowsheet, and attach a Pump (found under Pressure Changers item directory) to the absorber bottoms, then add another RateFrac column to the stream coming from the pump. The diagram on the next page is a visual description.

o Note: In the final design, it is probably a good idea to add a pressure changer to the Residual Gas Stream, which exits at 80 psig. Re- pressurizing your residual gas (or adding a multi-stage mixed-propellant cascade refrigeration unit) will cause it to condense, thus making your methane saleable as well as yielding a 600-fold decrease in volume and making it easy to transport .

decrease in volume and making it easy to transport . o o o o o Reopen

o

o

o

o

o

Reopen the browser and add the setup information to the Pump block. This will really only include the discharge pressure of 400 psig.

Now you need to set up the stripping column. This should be very similar to the manner in which you set up the absorber, except that you now have to include a Reboiler when you get to the Setup page. This also means that, if you have 10 segments, the Ending Segment is now 9.

o Note: Instead of using 10 segments, I recommend you try using 5 – 7. I speak from experience when I say that you will probably receive errors otherwise. I am not trying to dissuade you from using 10, merely saying that it happened when I ran my column.

You have to assign it an Operating Specification, such as distillate flow rate, boilup ratio, reboiler duty, etc, so that it can minimize the number of variables. I personally think that using the Distillate Flow Rate makes it easier to find out later on how large your recycle stream has to be in order to achieve your specifications. A good initial estimate for your distillate flow rate should be around the size of the molar flow rate of the amount of impurities you are trying to remove (Methane, Ethane, CO 2 , N 2 ), so try using a molar distillate rate of 1.75 lbmol/hr.

Again, here we don’t know the column pressure. Shown to the right is a sensitivity analysis plot for % of component retained in the bottoms vs. column pressure. According to the chart, the optimum pressure to operate at would be approximately 30 psi, since the minimum ethane composition occurs there.

% Component Retained vs Column Pressure for Initial Distillate Rate of 1.75

100% 90% 80% 70% 60% Ethane 50% Propane 40% Iso-Butane 30% 20% 10% 0% 0
100%
90%
80%
70%
60%
Ethane
50%
Propane
40%
Iso-Butane
30%
20%
10%
0%
0
100
200
300
400
% Component Retained

Column Pressure (psi)

Reinitialize, and run your simulation. Hopefully your results match ours below.

Page 33

While we do manage to remove a lot of the ethane bottoms stream, we also

While we do manage to remove a lot of the ethane bottoms stream, we also remove more propane than our spec allows. For right now, this is fine, since we will be recycling the stripper vapor back to the feed. So, now connect the STRIPVAP stream to the mixer, reinitialize and run the simulation. The results you obtain are ok, but still do not manage to meet the specifications that we were hoping for. To find the best separation possible, try running a Sensitivity Analysis comparing component molar flow rates to distillate rates. If you do, you should find that the best operating range for your distillate flow rate is between a flow rate of 1.99 to 2.4 lbmol/hr. Our diagram is shown to the right.

If you re-Run your simulation, you should obtain a bottoms stream fairly free of ethane that yields approximately 85% recovery. One thing to take note of is that the higher the distillate rate you use, the less room the depropanizer has to meet the recovery specification. While the value we used, D = 2.38, was on the higher end of this operating range, this was OK since our depropanizer does a very good job of separating out our propane from our butanes. The final results for our absorber and stripping column with recycle are shown below.

% Component Retained in Bottoms vs Molar Distillate Rate

100% 90% 80% 70% Methane 60% 50% Ethane Propane Iso-Butane 40% 30% 20% 10% 0%
100%
90%
80%
70%
Methane
60%
50%
Ethane
Propane
Iso-Butane
40%
30%
20%
10%
0%
% Component Retained

1

1.25

1.5

1.75

2

2.25

2.5

Molar Distillate Rate (lbmol/hr)

1.75 2 2.25 2.5 Molar Distillate Rate (lbmol/hr) To produce saleable propane and butanes, we’ll now

To produce saleable propane and butanes, we’ll now need to introduce and design in Aspen the two remaining columns: the DePropanizer and the DeButanizer.

Page 34

DePropanizer

o

The feed into this column will be the bottoms from the stripper:

 
 

Component

lb/hr

lbmol/hr

Mole %

Ethane

0.43

0.0144

0.14

Propane

49.16

1.1158

10.84

iso-Butane

17.46

.3004

2.92

n-Butane

35.00

.6022

5.85

iso-Pentane

4.35

.0603

0.59

n-Pentane

14.50

.2009

1.95

iso-Hexane

1.73

.0201

0.20

n-Hextane

10.39

0.1206

1.17

n-Heptane

6.04

0.0603

0.59

CO 2

0.017

.0004

0.004

n-Nonane

999.9

7.7960

75.75

Total

10.2914

100

o

You should be decently familiar with how to set up a distillation column in Aspen, so we won’t walk you through the basics. Refer back to Chapter 5 to navigate yourself through the simulation.

o

After inputting the feed rates, we’ll use the depropanizer in Chapter 5 as a start to simulating this column. We’ll start with 35 trays, feeding on tray 11, with a reflux ratio of 4.2 and distillate rate of 1.3 lbmol/hr.

o

While running this column, stages below the feed started to dry up, so we lowered the condenser pressure slowly until our column converged at 160 psia. Our first converging run produces the following results:

Our first converging run produces the following results: o o All of the propane is removed

o

o

All of the propane is removed out the tops, but the purity is not at 99% yet.

We tried lowering the distillate rate, and lowering the condenser pressure if the column dried up. While doing this, we maintained total recovery of propane, while increasing the purity of the tops to over 98.7%. We ended up with a condenser pressure of 140 psia (below).

We ended up with a condenser pressure of 140 psia (below). We cannot really improve this

We cannot really improve this separation, since ethane will come out with propane, and and the molar flowrate of ethane is 1.29% of that of propane, we won’t ever reach a 99% purity level in the tops.

Lastly, decreasing the size of the column to 30 trays and continuing to feed at tray 16 results in the same level of recovery and purity.

Debutanizer Next, we desire to recover as much of the butanes as possible from the bottoms of the depropanizer, with at least 97% purity.

o

We’ll start the column with 30 trays, feeding at tray 15, and setting our reflux ratio to 1.5. and distillate rate to 0.91 lbmol/hr. We set the condenser pressure to 100 psia, which is the vapor pressure of iso-Butane at 120 o F.

o

The results from our initial run are shown on the next page.

Page 35

o Here we see that we’ve recovered 100% of the butanes, and the purity of

o

Here we see that we’ve recovered 100% of the butanes, and the purity of the stream is over 99%. Excellent results for the first shot.

o

We’ll try to make the column a little smaller, by running a Sensitivity Analysis on reflux ratio, and feed tray for several column sizes.

o We were able to decrease the reflux ratio to 0.9, and use 25 total trays, while feeding on tray 9:

to 0.9, and use 25 total trays, while feeding on tray 9: Our final results for

Our final results for the debutanizer are almost 100% recovery of iso-Butane, 99.8% recovery of n-butane, and a stream purity of 99%.

When you are modeling your process and you find that nothing seems to work, just try making slight adjustments, perhaps in error tolerance, or increment size. Mastery of ASPEN comes in time, but also understand that small things like that have, in reality, major consequences on how ASPEN operates. Fiddling with, or, as I like to call it, “fine-tuning,” can mean the difference in getting severe ASPEN errors on the first iteration and seeing your 7 column system with 3 recycles converge in 4 iterations. If you are still stuck, possibly consider modifying your process slightly. This could include adding a small purge stream, to aid in convergence, or specifying the data associated with a stream.

Wow, wasn’t that fun? Sorry for the blatant sarcasm (we thought it would introduce an element of levity), but I am sure you will have as much fun with this example as we did. Recycle streams are easily the worst possible curveballs you will have thrown at you, and to conquer that is seriously quite an accomplishment. We have only worked on a simple case here, but, much like all chemical engineering courses, it is our wish that you are able to adapt what you have learned here to your personal situation.

Now that you, hopefully, have overcome the challenge of Aspen, you will need to create a project cost estimate with the help of Aspen Icarus Process Evaluator.

Page 36

Chapter 8: Icarus Process Evaluator (IPE) Mapping ISBL and OSBL

Once you’ve got your process design down, your manager will probably want to know something about cost. Aspen Icarus provides a more precise cost estimation than that of the FACT Method, and can be used to price everything from your columns, reactors, and other process equipment (ISBL), to the pavement on your parking lot, and fire hydrants on the road (OSBL). It works in a similar fashion as FACT method, by pro-rating each item and area, but does so in a more detailed manner.

In evaluating your process, you’re going to need to report ISBL and OSBL. ISBL (Inside Battery Limit) facilities are comprised of major purchased equipment costs, and the costs associated with that equipment. OSBL (Outside Battery Limit) facilities include storage tanks, shipping for feedstock and product, cooling water, steam, fire water, instrument air, flare stack, environmental treating, and other facilities (FACT description). Appraising ISBL and OSBL is done through Aspen Icarus Project Evaluator.

Before you begin, the first thing you will need to do is to fix your simulation so that it will import correctly into Icarus. It’s not the simplest thing to remember, but it really should be done to ensure your Aspen file will be properly detailed and analyzed by Icarus.

Fixing an Aspen Simulation to Work in Icarus:

o

With your Aspen Simulation open, go to File Import, and first change the file type to Templates (*.apt). You will need to open a file located under C:/Program Files/Aspen Tech/Aspen Plus 11.1/GUI/ Templates/Simulations/ called Aspen IPE Stream Properties.apt.

o

Now go to your Data Browser, open the Components folder, and then open the Specifications subfolder. Click on the Databanks tab, and highlight all of the fields on the right.

o

Click the

o Click the button. Reinitialize and Run . This should allow you to properly import your

button. Reinitialize and Run. This should allow you to properly import your working Aspen files. Now you will need

to send your project to Icarus and map your equipment.

Mapping ISBL:

o

o

o

o

o

o

o

o

o

o

o

Once you have your Aspen Plus file fixed to import, go to File Send to Aspen Icarus. On the left-hand side of the screen, you should see the following tabs:

side of the screen, you should see the following tabs: Select the first tab, Project Basis

Select the first tab, Project Basis View. In the Basis for Capital Costs folder, click on General Specs .

There is a field labeled Project Type, which indicates what type of a project this is (addition/new). For a plant addition (i.e. for U.O. Labs) change Grass Roots/Clear Field to Inside Plant Addition.

Now, select the center tab, Process View, and left click your design. Select Map to turn it into an Icarus component. You may also click

on

For columns, when you reach the Project Component Map Preview screen, check out the Configuration option box. This lists ten different configurations (use the arrows to scroll up and down – it only looks like it has two options) to choose from. The detailed descriptions are on page 4-30 of the Icarus manual. o Important Note: Make sure you choose a configuration that does NOT have a split condenser or you will run into problems later on.

You can also add extra equipment to the blocks you are mapping by selecting New Mapping. Click OK to map.

To add extra pumps or peripheral equipment, on the left-hand side of the screen, click on the third tab, Project View. Right-click on the main process area, and Add Project Component. The equipment that is of concern to you at this point is listed under Process Equipment.

Once added, right-click on each item and click Size.

Enter any required input; this includes known flow rates, pressure head required, etc.

Click on

Select the area you are dealing with, and on the right side, select List. It should look something like the image to the right.

List . It should look something like the image to the right. to map all items.

to map all items.

something like the image to the right. to map all items. (or Run ‡ Evaluat ion
something like the image to the right. to map all items. (or Run ‡ Evaluat ion

(or Run Evaluation Project) and Evaluate All Items.

o

o

o

Right-click on each item and choose Size to view the sizing information Icarus obtained from Aspen.

You can also right-click on each item and choose Item Report to view the sizing and cost data. You can also select multiple items at once to view multiple reports. Go ahead and fix anything that doesn’t look right.

Clicking the Results tab will reveal your overall capital and utilities costs.

Page 37

Once you’re confident that you have correctly mapped your ISBL components, you can move on to OSBL. A lot of OSBL tends to be project- specific, so be careful with sizing. Most importantly, make sure whatever you tell Icarus, as well as what Icarus tells you, makes sense. Mapping OSBL:

Icarus is a good tool for estimating the cost of OSBL within your facility. You have probably already used the FACT method to calculate costs of OSBL, and you might find that the numbers are fairly similar. Understand that although Icarus may be more exact with their calculations

(FACT only uses percentages of one cost to estimate another), it is also just an estimation based on sizing. For example, it will estimate the size of water facilities, or amount of wiring, based on the area of your plant. Before adding OSBL facilities, note that the following elements are sized with a default value, and will need to be resized:

Utilities (except for Electricity)

o These are read from the PROJSUM.ICS reports file. Within the Results tab click on the Project Summary tab (it may be hidden if the middle window isn’t expanded enough – its tab should be next to Executive Summary):

enough – its tab should be next to Executive Summary ): Area Sizing/Pipe lengths : o

Area Sizing/Pipe lengths:

o

Area sizing is needed for infrastructure sizing, and pipe lengths are needed for flare headers, process lines, steam lines, condensate returns, etc. o Note: To proceed with these sizes, adjust your VISIO diagram to scale for the now known equipment sizes.

o

When your scale is under control, you should be able to read area sizes and your pipe rack with a ruler. In Icarus, Areas refer to objects that contain particular types of mapped equipment, not necessarily areas of your plant.

o

Piping and wiring lengths are calculated from the center of an Area so be sure that the sizing and type of area specified is accurate.

o

The number and size of each type of storage tank should have been calculated previously.

o

The wastewater treatment and settling pond sizes are also calculated depending on what your process requires.

What you do NOT need to add:

o On the left -hand side of the screen, click on the Project Basis tab. In the Basis for Capital Costs folder, click on General Specs . There is a field labeled Project Type, where you can indicate if this is a type of plant addition or a brand-new project. You should choose Grass roots/Clear field, which indicates that this is a new project. Icarus automatically adds certain components, which you do not have any control over, so do NOT add these components:

   

Components Included

Project Type

MAIN Substation

UNIT

Control

Grass roots/Clear field

Transformers,

MCC*, SW

Operator Center,

Switchgears

Transformer

Control Center

Page 38

Adding OSBL:

o On right side of screen, click the Projects tab and the + sign next
o
On right side of screen, click the Projects tab and the + sign next to Additional Project Components.
o
Double-click on either inch-pound or metric and a window similar to the one shown below should open in the right-hand window:
o Note: Each
is considered an “Area” by Icarus. You can either drag the entire area to your project in the left-hand window, or
expand the area and drag certain components.
o
You might want to create/drag areas for storage, wastewater, main process, and plant area (as big as the whole plant for parking,
buildings, roads, etc).
o
Once you’ve dragged over the components you need, double-click on them in the left -hand window (Project View) to see their default
sizes, and make necessary adjustments to their size or Number of identical units (number you wish to purchase). Check the cost, because
some items are too project-specific for IPE to calculate, and you may need to research a cost and add it in.

Some OSBL facilities to consider:

Water Raw water treating

(Make up for BFW + process water + cooling tower) x 1.1

De-Aerator

Clarification and/or silica removal may be required if city water is not the raw water source. Boiler feed water (BFW comprises all steam demand – non-recovered condensate or process steam) x 1.1. Demineralization required All water going to steam generators

Electric Power Incoming power switch yard

Total electric power load

Motor control centers.

Un-interruptible Power Emergency Generator

All electric motors in facility. One or more for process And one or more for OSBL Sufficient for all instrument and shut down systems and emergency lighting (say 5 kW minimum) Say 50 kW minimum

Air System (Instrument Air and Yard Air) Typically air is available to the consumer at 90 psig. Suggest minimum 300 scfm

Compressor(s) Air Dryer Air Receiver

Usually two at 100%; one motor and one turbine drive Usually one sized for 100% + spare 5 minutes of capacity going from 90 to 60 psig.

Fuel Gas

Incoming gas metering station

Fuel gas supply drum

Fuel gas collection drum if process makes useable fuel gas

(5 minutes of capacity going from 60 psig to 45 psig with main supply source not flowing)

Page 39

Fire Fighting Facilities Central Fire Pumps

Central Fire Water Tank Ring Mains Hydrants and monitors Breathing Apparatus Fire Truck Foam inventory Fire Station

200-psig discharge pressure minimum. 1 motor, 1 turbine and 1 diesel minimum) Eight hours at maximum capacity of one fire pump Size for 1500 gpm flow or 8 inch diameter minimum Sufficient to attack the fire from at least two locations

(Should have some sort of fire fighting truck. May need full size engine)

Aqueous Effluent Treating Plant This is very much a function of the plant and the types of aqueous effluents. As a minimum, require a collection tank, and aerated bio-digester followed by polishing tank or pond and land farm for sludge disposal. If you can hook up to the municipal sanitary sewer, so much the better, if not then you need a separate sanitary waste treating

Flare System Flare line (s) Flare Drum (s) Flare Stack (s) (minimum height 150 feet) Flare drum pump out pumps

Roads, Bridges, Fences, Parking Lots Should be around 1 – 2 % of total installed cost dependent upon plant size

Raw Material Receipt and Product Shipping Rail siding Truck loading bays Pipeline Connection Warehouse

Maintenance Equipment Cranes Trucks Other Vehicles Welding Machines Other Special Equipment Hand Tools

Buildings Central Control Building\ Product Loading System Control Building Dispatcher’s Office Gate House First Aid Fire House Admin Building Maintenance Machine Shop Spare Parts Store House

These services can be combined in one or more buildings depending upon plant size

Spare Parts Commissioning and Start Up

Insurance Spares

1% of all equipment and materials Percent of equipment capital cost Fired Heaters @ 10% Heat Exchangers @ 4 % Pumps and Compressors @15% Columns, Vessels, Reactors @ 1%

Page 40

Invested Spares

Fire and Safety Equipment @ 14% Instruments @ 10% Electrical @ 13% 25% of major critical un-spared equipment such as compressors and high pressure pumps

Storage Tanks The sizing of storage tanks is a function of the amount of material that is likely to flow in or out of them in a given time period.

Feedstock. If a storage tank is expected to receive one lot of feedstock equivalent to 20 rail cars of 25,000 gallons capacity per car, then the tank must be able to receive at least 500,000 gallons of material. Since it would make no sense to run the tank totally empty or totally full, tanks are specified for a "working capacity. This "working capacity" is then multiplied by say 120% to give the actual size of the tank. So for 500,000 gallons of working capacity we would specify a 600,000-gallon tank for this service.

Continuing this example, the plant would have at least two 600,000-gallon tanks in the feed system. One tank would be receiving feedstock, and the other would be providing feedstock to the process. If the feedstock delivery method is somewhat erratic, or it more or less tank cars might make up a shipment, then there may be need for further tankage.

For added flexibility you might also choose to have three 400,000-gallon tanks rather than two 600,000-gallon tanks.

All incoming feedstock is usually tested for quality assurance purposes, either by random samples or by testing a tank of received product.

If your feedstock is being delivered by pipeline then it is common to run to a tank and then into the process. You would most likely fill the tank one-day, test it the next and release it for processing on the third. This suggests that you should have three tanks each with say 120% of one day's capacity.

Product Product storage is a function of testing policy and shipment size.

It is industry practice to run a full lab test on a tank before it is released for shipment. Typically samples are taken from the top, middle and bottom of the tank and are tested both individually and as a composite sample. That means that the tank must have no flow in or out from the time the samples are taken. Any product produced in that period must go to another tank(s).

There must be enough tankage for supplying the maximum sized product shipment, plus testing a full tank, plus rundown capacity for product.

In lieu of a witty comment, look at this amusing Dilbert anecdote with the full knowledge that we are nearing the end.

of a witty comment, look at this amusing Dilbert anecdote with the full knowledge that we

Page 41

Chapter 9: Aspen Pinch An Introduction to Aspen Pinch

While tackling the issue of Heat Integration throughout Design, your main objective is to arrange your process to provide the most economically efficient heat transfer. Once you’ve optimized your design of reactors, separation units and recycles, you have a fixed system of material and energy balances from which you can design a heat exchanger network. Then cooling and heating requirements left over from this scheme must be satisfied by utilities.

When designing a heat integration scheme, you should balance two economic factors: operating cost and capital cost. Operating cost is calculated from minimizing utilities, and capital cost is calculated from minimizing the number of heat exchangers. Though you may design an intricate heat exchanger network that minimizes utilities, your capital costs will skyrocket purchasing all those exchangers. Alternatively, with few exchangers and low capital cost, you will have a lot of utility to make up for, and thus your operating cost will soar. Pinch Technology calculations assist the process engineer in achieving the balance by choosing the best streams matches for streams to exchange.

Pinch Technology uses thermal data from streams involved in heat integration to assist in designing an efficient network. The first step in heat integration is to define hot and cold streams. “Hot” streams are streams that need to be cooled, and “Cold” streams are streams that need to be heated. Let’s step back and work through a simple example.

Problem Statement:

Consider the following scheme, with three columns, a reactor, and a mixer and use Aspen Pinch to optimize your heat integration:

mixer and use Aspen Pinch to optimize your heat integration: Solution Methodology: o First, we’ll develop

Solution Methodology:

o First, we’ll develop a stream table that will display the data needed to solve the heat integration problem. o Note: The names for streams that you choose in your flowsheet are not

going to be the same names that Aspen Pinch chooses. We can easily see that one hot stream is the Reactor Effluent, and two cold streams are B1 Feed and B2 Bottoms.

o

In addition, utility is also consumed for each reboiler and condenser, so we must include those items in our stream table.

o

As you might remember from Chapter 3, you can obtain the heat capacity of each element at various temperatures by doing a Property Analysis. For each stream, we used a weighted average of each stream’s C p at T 1 and T 2 , then averaged the two for an average heat capacity.

o

The enthalpy of each stream also can be taken from Aspen, subtracting the enthalpy of each stream at its initial temperature from its final temperature.

o

Our completed stream table is shown to the right.

Stream

Type

T1

T2

dH

 

(

o F)

(

o F)

(MMBtu/Hr)

R1 Effluent

HOT

200

140

-4.12

B1 Feed

COLD

100

145

1.46

B2 Bottoms

COLD

160

200

4.32

B1 Condenser

HOT

135

135

-3.3

B1 Reboiler

COLD

150

150

5.8

B2 Condenser

HOT

140

140

-28.4

B2 Reboiler

COLD

160

160

28.4

B3 Condenser

HOT

140

140

-39.7

B3 Reboiler

COLD

170

170

35.6