Distillation design and control using aspen simulation download free

The purities are too low, so we need to increase the reflux ratio or add more stages to get a better separation. If we go back to Setup in the C1 block, change the reflux ratio to 3, and click the N button, the simulation converges with the results shown in Figure 3.

So we are getting pretty close to the desired purities. We could continue to manually change the reflux ratio and the distillate flowrate to attempt to achieve the desired product purities by trial and error. However, there is a much easier way, as discussed in the next section. To achieve these precise specifications, Aspen Plus uses the Figure 3. The simulation attempts to adjust the manipulated variable in such a way that the specified value of the controlled variable is achieved.

A word of caution might be useful at this point. The solution of a large set of simul- taneous nonlinear algebraic equations is very difficult. There is no guarantee that a solution will be found because of numerical problems. In addition, if good engineering judgment is not used in selecting the target values, there may be no physically realizable solution.

For example, if the specified number of stages is less than the minimum required for the speci- fied separation, there is no value of the adjusted variable that can produce the desired result. Another possible complication is multiplicity. Because the equations are nonlinear, there may be multiple solutions. Sometimes the program will converge to one solution and at other times it will converge to another solution, depending on the initial conditions.

It is usually a good idea to start by converging only one variable at a time instead of trying to handle several simultaneously. In our example, we will converge the distillate specifica- tion first by adjusting distillate flowrate. Then, with this specification active, we will con- verge the bottoms specification by adjusting the reflux ratio. The order of this sequential approach is deliberately selected to use the distillate first because the effect of distillate flow- rate on compositions throughout the column is much larger than the effect of the reflux ratio.

Clicking the New button opens the window shown in Figure 3. On the first one, Specifications, you can specify the type of variable and what its desired value is. Clicking the dropdown menu under Design specification and Type opens a long list of possible types of specifications Fig. Select mole purity. Then click the second page tab Com- ponents. Click the IC4 in the left column under Available components.

The Design Spec is now completed. Now we must specify what variable to adjust. Clicking the Vary item under the C1 block opens the window shown in Figure 3. Opening the dropdown menu under Adjusted variable and Type produces a long list of possible variables.

We select Distillate rate, which opens several boxes Fig. We set the lower bound at 0. Note that all the items in the Data Browser window are blue, so the simulation is ready to run.

We click the blue N button and run the program. The Control Panel window opens and tells us that it has taken three iterations to converge Fig. Going down to Stream Results at the bottom of the list under the C1 block lets us look at the new values of the stream properties.

Note that the flowrate of D1 has changed to 0. Then the mole purity of the bottoms B1 is specified to be 0. See Figures 3. Clicking the blue N button executes the program.

The simulation converges in three iterations. The mole fraction of iC4 in the distillate is 0. Both are now very close to their specified values. Of course, the distillate flowrate and the reflux ratio have Figure 3. The stream results show that the flowrate of D1 is 0. To ascertain the reflux ratio, click on Results Summary under the C1 block. The reflux ratio is 3. The other important pieces of information in the window are the condenser heat removal [ To attain the desired K, the pressure should be increased a little.

If we rerun the simulation with a pressure of Of course, at this new pressure the required reflux ratio changes. It increases from 3. This shows the adverse effect of pressure on relative volatilities that occurs in most hydrocarbon systems. The column should be operated at as low a pressure as possible to save energy.

The most important piece of information is the reboiler heat input The base temperature is This will dictate the pressure Figure 3. The stream conditions at the The column temperature and composition profiles can be obtained by selecting Profiles in the C1 block.

The window that opens is shown in Figure 3. The first TPFQ temperature, pressure, flow, heat gives the temperature and pressure on each stage. Selecting the second page tab Compositions opens the window shown in Figure 3. Using the Plot Wizard program makes generating plots of these profiles quite easy.

Click on Plot at the top toolbar of the Aspen Plus simulation window. Then click Plot Wizard. Clicking Next opens the window shown in Figure 3. Clicking on the upper left picture labeled Temp produces the temperature profile plot given in Figure 3. Clicking on the picture labeled Comp and then clicking Next opens the window on which you can select what components to plot and what phase liquid or vapor compositions.

In addition, the minimum reflux ratio and the minimum number of trays can be determined. These will be useful for heur- istic optimization, which is discussed in detail in Chapter 4. Of course, if refrigeration were used in the condenser, this heat removal expense would also be quite large. Therefore, reboiler heat input is the quantity that should be minimized. The simulation is run using different feed stages. The purities of both products are held constant. The feed stage that minimizes reboiler heat input is the optimum.

Table 3. Feeding on stage 14 gives the minimum energy consumption. TABLE 3. It is assumed that the feed stage is a fixed ratio of the total number of stages. The results, given in Table 3. Product purities are held constant. Results are given in Table 3. The typical distance between trays tray spacing is 0. If there are NT stages, the number of trays is NT 2 2 one stage for the reflux drum and one for the reboiler. In addition to the trays, some space is needed at the top where the reflux piping enters the vessel and at the feed tray for feed distribution piping.

More significantly, space is needed at the base to satisfy two requirements: 1 liquid holdup is needed for surge capacity and 2 the liquid height in the base of the column must be high enough above the elevation of the bottoms pump to provide the necessary net positive suction head NPSH requirements for this pump.

So the length of the vessel can be estimated from the following equation. If this velocity is exceeded, the column liquid and vapor hydraulics will fail and the column will flood.

Reliable correlations are available to determine this maximum vapor velocity. Since the vapor flowrates change from tray to tray in a nonequimolal overflow system, the tray with the highest vapor velocity will set the minimum column diameter.

If the vapor mass flowrate and the vapor density are known, the volumetric flowrate of the vapor can be calculated. Then, if the maximum allowable velocity is known, the cross- sectional area of the column can be calculated. Aspen Plus has an easy-to-use tray sizing capability. A window will open, where the column sections to be sized and the type of tray can be entered.

The stages run from stage 2 the top tray to stage 31 the bottom tray. Sieve trays are specified. The simulation must be run by clicking the N button. Then the page tab Results is clicked see Fig. This is a very large distillation column, and therefore a single liquid pass would produce very large liquid gradients across the tray and liquid heights over the weir.

A column this large would use at least 2-pass trays. Changing the number of passes to 2 on the Specifications page tab produces a large change in the calculated diameter, dropping it from 7. All the detailed information about the vapor and liquid flows throughout the column can be accessed by clicking the subitem Report under the C1 block and under Property Options checking the box in front of Include hydraulic parameters.

Then, after the program is run, click the subitem Profiles and click the Hydraulics page tab. The window that opens gives lots of information about liquid and vapor rates and properties, as shown in Figure 3. The maximum vapor volumetric flowrate is 9. The vapor density on this stage is 2.

Using an F factor of 1, the maximum velocity is 0. This corresponds to a diameter of In either case, this is a very large distillation column, and, as we will see in the next chapter, it is very expensive to buy.

These methods are applied to a variety of columns in later chapters. The calculated design par- ameters included the operating pressure of the column, the reboiler and condenser heat duties, and the length and diameter of the column vessel. In this chapter the steady-state economic optimization of a distillation column is dis- cussed. Basically, we need to find the optimum number of total stages. There are some simple approaches, and there are more rigorous approaches. The simple methods use heuristics such as setting the total number of trays equal to twice the minimum.

The rigorous methods determine how the capital and energy costs change with the number of trays and find the minimum total annual cost design.

We will discuss both of them in this section and compare the designs that result from applying each. It should be emphasized that they cannot both be used simultaneously for rigorous design because fixing one of the two completely specifies the design of the column. We found the minimum number of trays more rigorously in Chapter 3 by using the simulator to find the number of stages where the required reflux ratio became very large.

Taking twice this number and adding two stages for the reflux drum and reboiler give a stage column, which is the column we designed in Chapter 3. It is interesting to compare this rigorous number with what the Fenske equation predicts for the same system. The usual approach is to find the relative volatility at the temperature at the top of the column stage 2 temperature is By definition, relative volatility aLH is the ratio of the vapor and liquid compositions of component L divided by the same ratio of component H.

This would give 32 stages. The minimum reflux ratio found in Chapter 3 from the simulation was 2. Multiplying this by 1. This is very close to the reflux ratio 3. It is interesting to compare the minimum reflux ratio found in the simulation with that predicted by the Underwood equations. These equations are derived assuming constant relative volatilities.

As we have seen in the previous section, the relative volatility between propane and isobutane is almost constant. It varies from 1. Therefore, the Underwood equations should predict the minimum reflux ratio quite well. As discussed in Chapter 2, there are two equations. Smaller items such as pumps, valves, and the reflux drum are seldom significant at the conceptual design stage. The cost of the trays themselves is usually small compared to the costs of the vessel and heat exchangers unless expensive internals are used such as structured packing.

Table 4. The sizing of the column vessel has been discussed in Chapter 3. The condenser and reboiler heat duties are determined in the simulation, but we need to have an overall heat transfer coefficient and a differential temperature driving force in each heat exchanger to be able to calculate the required area. The values of these parameters given in Table 4. Note that the overall heat transfer coefficient of the condenser is larger than that of the reboiler.

Reboilers have a greater tendency to foul because of the higher temperature more coking or polymeriz- ation and because any heavy material in the feed drops to the bottom of the column. Various objective functions are used for economic optimization. However, many assumptions must be made in applying these methods, and the accuracy of these assumptions is usually quite limited.

The predic- tion of future sales, prices of raw materials and products, and construction schedule is usually a guessing game made by marketing and business managers whose track record for predicting the future is almost as poor as that of the weather forecaster meteorologist. Therefore, the use of some simple economic objective function usually serves the purpose of optimizing a distillation column design.

We will use the total annual cost TAC. As shown in Table 4. The units of capital investment are U. TABLE 4. In some locations energy sources are plentiful and inexpensive. For example, in Saudi Arabia gas coming from an oil well is sometimes simply flared burned. In other locations, fuel is quite expensive because it must be transported long distances. For example, in Japan some of the natural gas is shipped in from Indonesia on liquefied natural gas tankers LNG , which are very expensive.

Therefore energy costs depend on location. The numerical example is for the stage column studied in Chapter 3. The stage case is shown in the second column. The capital cost of the column shell, which is 5. The other columns in Table 4. If the number of stages is reduced to 24, which gives a shorter column, reboiler heat input increases.

This increases column diameter and heat exchanger areas. This results in an increase in both capital and energy costs. If the number of stages is increased, the column becomes taller, but its diameter becomes smaller because reboiler heat input decreases. This decreases heat exchanger costs and energy costs. However, the cost of the vessel increases because it is longer. So the effect of increasing the number of stages is to increase the capital cost of the shell and to decrease the capital cost of the heat exchangers and energy costs.

The cost of the shell continues to increase to the 0. Figure 4. These differences may seem quite large and indicate that the heuristics are not very good.

However, good engineers always build in some safety factors in their designs. The number of trays in a column can sometimes be increased by going to smaller tray spacing or installing more efficient con- tacting devices. But changing the diameter requires a completely new vessel. Therefore, the heuristics give a pretty good design. It should also be noted that the optimum is quite flat. The TAC decreases only from 5. If the cost of energy is reduced, the optimum number of stages becomes smaller.

It is clear that energy costs dominate the design of distillation columns. Stainless steel is used in the cost estimates given in Table 4. If the materials of con- struction were more exotic, the optimum number of stages would decrease.

There are several types of rating problems. One of the most common is finding the product purities that maximize profit. In the design problem considered in previous sec- tions, we assume that the product purities were given.

In many columns the purity of one product may be fixed by a maximum impurity specification, but the other product may have no set purity. We know that distillate flowrate should be maximized and that as much isobutane as possible should be included in this stream, up the impurity constraint.

This can be achieved by minimizing the concentration of propane that is lost in the bottoms. But reducing xB requires an increase in reboiler heat input, which increases energy cost. Therefore, there is some value of xB that maximizes profit. The optimization must take into account the value of the propane product compared to the bottoms and the cost of energy.

The steady-state simulator can be used to find this optimum operating condition. Then a new value of bottoms composition is specified and the calculations are repeated. There is a rapid rise in reboiler heat input below 0. The maximum profit is obtained with a bottoms composition of 0. The window shown in Figure 4. On the Define page the variables to be used in calculating the profit are defined.

Type a variable name under the Flowsheet label. Reboiler heat input is QR in watts. Placing the cursor on one of the lines and clicking the Edit button open the windows shown in Figure 4.

For example, FW is edited in Figure 4. Under the Category heading, Streams is selected. Since it is in the C1 block, Blocks under the Category heading is selected. This variable is defined by clicking the Fortran page tab and entering the equation for profit as shown in Figure 4. Selecting the final page tab Vary opens the window shown in Figure 4. The variable selected to vary in order to find the maximum profit is the reflux ratio. When the window first opens, the box to the right of Variable number is blank.

Right-clicking opens a little window with Create that can be selected. The optimizer is now ready to run. Clicking the N button executes the program.

The Control Panel window Fig. The optimum value of bottoms composition is 0. The reboiler heat input is The approaches presented are simple and practical. There are many advanced techniques in the optimization area that are beyond the scope of this book. This is a fairly ideal system from the standpoint of vapor — liquid equilibrium, and it has only two components, a single feed and two product streams. In this chapter we will show that the steady-state simulation methods can be extended to multicomponent nonideal systems and to more complex column configurations.

Other methods of analysis can also be applied to these more complex systems. In par- ticular, we will find that the use of ternary diagrams provides very useful insight into the design of these complex systems. In addition, the effects of various design parameters in these nonideal systems are sometimes counterintuitive and significantly different from those in an ideal system. Mixtures of methyl acetate, metha- nol, and water are generated in the production of polyvinyl alcohol.

Both the methyl acetate and the methanol must be recovered for recycling or for further processing. This means that the overhead product from a distillation column separating this binary mixture cannot have a methyl acetate composition greater than this azeotropic composition. As discussed in Chapter 1, Aspen Plus has several nice analysis tools. The window shown in Figure 5. The dropdown menu at the top left gives three choices. Since distillation columns run at fixed pressures, the Txy diagram is the most appropriate.

We set the pressure at 1. Note that the Wilson physical property package has been specified. Clicking the Go button at the bottom of the window produces the Txy diagram shown in Figure 5. The minimum- boiling homogeneous azeotrope at Figure 5. Figures 5. The pure component boiling points at 1. Methyl acetate is Vapor —liquid equilibrium information for the ternary system can be seen by going up to the top toolbar and clicking Tools, Analysis, Properties, and Residue. Clicking the Go button at the bottom of the window produces an equilateral ternary Figure 5.

The residue curve lines lead from the minimum-boiling azeotrope to the highest-boiling component, which is water. Note the distillation boundary that runs between the two azeotropes. As discussed in Chapter 1, Aspen Split provides additional insight. Click on Tools, Analysis, and Aspen Split.

Two choices can be made: Azeotropic Search and Ternary Dia- grams. Selecting the first opens the window shown in Figure 5. Clicking on Azeotropes in the list on the far left of the window under Output produces the results shown in Figure 5. Clicking Report gives infor- mation about the two azeotropes Fig. The second alternative in Aspen Split is to select Ternary Diagrams. This opens the window shown in Figure 5.

Then clicking Ternary Plot at the bottom of the list on the left side of the window opens a right triangle, which displays boiling points of the pure components, the compositions and temperatures of the azeotropes, and the distillation boundaries see Fig.

To add the residue curve lines, right-click on the graph, select Add and Curve. Move it to a spot on the diagram and left-click. A residue curve through that point is drawn. Note that a grid has also been added. This is achieved by clicking the bottom icon on the far right that is Toggle Grid. A number of parameters of the plot can be changed by right-clicking and selecting Properties. The color and thickness of lines and curves can be specified, and the plot can be rotated if desired.

Another useful feature is the ability to add a marker at some desired location. This is done by clicking the sixth icon from the top on the right side of the window. A window opens on which you can enter the composition coordinates at which you want to place a mark. In Figure 5. Then click OK to place the mark at the correct location Fig. These markers can be used to show the location of the feed F , distillate D , and bottoms B , as illustrated in Figure 5.

Note that both the D and B points lie inside the upper region of the diagram and are not in two regions separated by a distillation boundary. Also note that these points lie near a residue curve.

Therefore this should be a feasible design. The concentration of the methyl acetate in the distillate should be fairly close to the azeotropic composition because this stream is sent for further processing.

In other plants the stream may be fed to a reactor where it reacts with water to form acetic acid and methanol. In any event, we want as high a concentration of methyl acetate as possible in the distillate. The concentration of water in the distillate should be 0. The concentration of methyl acetate in the bottoms should be 0. The feed flowrate is 0. Column pressure is initially set at 1. For a preliminary design, a column with 32 stages is selected.

We will vary the feed tray location to minimize reboiler heat input. Initially stage 16 in the middle of the column is specified.

Because of the methyl acetate azeotrope, the distillate cannot be richer in methyl acetate than The resulting product compositions are 2. To drive the bottoms methyl acetate compo- sition to the specified 0. The column does not converge after the default 25 itera- tions. Clicking Convergence under the C1 block lets us change the number of iterations to 50 and try again.

The column still does not converge. The problem is that we have been using the standard method of convergence. This must be changed to either Azeotropic or Strongly nonideal liquid to achieve convergence in this nonideal system. This is done on the Configuration page tab under Setup, as shown in Figure 5. It is sometimes necessary to switch back and forth between Azeotropic and Strongly nonideal liquid to achieve convergence. This is done by right-clicking on the individual Design Spec and selecting Hide.

The same pro- cedure is followed for the corresponding Vary. After the simulation has converged without the Design Spec and Vary active, they can be reactivated by right-clicking Design Spec or Vary and selecting Reveal. The new distillate flowrate to achieve the bottoms methyl acetate of 0.

The water composition of the distillate is now 0. The simulation is run, and the required reflux ratio is 4. Now the effect of the feed tray location is explored to find the optimum where reboiler heat input is at a minimum. However, we also need to keep track of what is happening to the methyl acetate concentration in the distillate. Unlike the ideal system explored in Chapter 3, energy consumption and product compositions are very sensitive to feed tray location.

The feed stage that mini- mizes reboiler heat input is stage 27, which is very near the bottom of the column. However, the methyl acetate composition in the distillate drops drastically for feed stages greater than stage Therefore a design tradeoff must be made between the cost of energy in this column and the cost of feeding a lower-purity methyl acetate stream Figure 5.

This is a good example of the typical multiunit tradeoffs that occur in complex chemical plants. This also illustrates that the simple heuristics cannot always be applied in complex nonideal systems.

The basic complexity in this example is that we have three design spe- cifications instead of the normal two. Therefore another degree of freedom is required, which is the feed tray location. The next step is to see the effect of using different numbers of total stages. The breakpoints where the distillate methyl acetate purity drops precipitously in the four cases are stages 12, 17, 21, and 26, respectively. At these four points, the reboiler heat inputs are 6.

So adding more stages reduces energy consumption up to about 27 stages and then it actually increases. This indicates that either the stage design or the stage design is the economic optimum. Evaluating the total annual cost of these four designs shows that the stage design gives the lowest TAC. Table 5. It is interesting to check the rigorous simulation developed above with the ternary analysis using the graphical column design method in DISTIL. The reflux ratio is specified to be 1, which would correspond to that found in the optimum stage design.

Note that the latter is close to the azeotropic composition. The total number of stages is 30 with the feed introduced on stage These results are in good agreement with the rigorous simulations. It should be noted that the results are very sensitive to the methyl acetate spe- cification in the distillate. TABLE 5. The components involved are ethanol, water, and benzene. The flowsheet contains two distillation columns and a decanter.

There are two recycle streams, which create very difficult convergence problems, as we will see. So this complex example is a challenging simulation case.

The origins of the example go back over a century when a process to produce high-purity ethanol was needed. Ethanol is widely produced in fermentation processes. This binary stream is then fed into a second distillation column.

The overhead vapor is a ternary mixture of all three components. When it is condensed, the repulsion between the water molecules and the organic benzene molecules is so great that two liquid phases form. The reflux drum becomes a decanter. The lighter organic liquid phase is pumped back to the column as organic reflux. The heavier aqueous phase contains significant amounts of ethanol and benzene, so it is fed to a second distilla- tion column in which the water is removed in the bottoms stream.

The distillate stream is recycled back to the first column. The first item to explore is the complex VLLE of this heterogeneous vapor — liquid — liquid system.

Two binary Txy diagrams are given in Figures 5. The two homo- geneous minimum-boiling azeotropes are clearly shown. The ternary diagram Fig. A large part of the com- position space has two liquid phases. The liquid —liquid equilibrium tielines are shown. The aqueous phase is on the left, and the organic phase is on the right. Note that the ternary azeotrope is heterogeneous and has the lowest boiling point This means that the over- head vapor from the first column will have a composition that is similar to that of this azeo- trope.

Note also that the diagram is split up into three regions by the distillation boundaries that connect the four azeotropes. As you may recall from Chapter 2, these boundaries limit the separation that is attainable in a single column.

A comprehensive and example oriented text for the study of chemical process design and simulation Chemical Process Design and Simulation is an accessible guide that offers information on the most important principles of chemical engineering design and includes illustrative examples of their application that uses simulation software.

A comprehensive and. This book offers a comprehensive coverage of process simulation and flowsheeting, useful for undergraduate students of Chemical Engineering and Process Engineering as theoretical and practical support in Process Design, Process Simulation, Process Engineering, Plant Design, and Process Control courses. The main concepts related to process simulation and application tools are. Distillation has historically been the main method forseparating mixtures in the chemical process industry.

However,despite the flexibility and widespread use of distillationprocesses, they still remain extremely energy inefficient. Increased optimization and novel distillation concepts can deliversubstantial benefits, not just in terms of significantly lowerenergy use, but also in reducing. The use of control systems is necessary for safe and optimal operation of industrial processes in the presence of inevitable disturbances and uncertainties. Plant-wide control PWC involves the systems and strategies required to control an entire chemical plant consisting of many interacting unit operations.

Over the past 30 years, many tools. Distillation design and control using aspen tm simulation by William L. Distillation Design by Henry Z. Kister,Henry Z.. Batch Distillation by Urmila Diwekar. Most importantly, as renewable sources of energy and chemical feedstocks continue to be developed, distillation design and control will become ever more important in our ability to ensure global sustainability.

Moreover, readers will discover how to develop effective control structures. While traditional distillation texts focus on the steady-state economic aspects of distillation design, this text also addresses such issues as dynamic performance in the face of disturbances.

Distillation Design and Control Using Aspen Simulation introduces the current status and future implications of this vital technology from the perspectives of steady-state design and dynamics. The book begins with a discussion of vapor-liquid phase equilibrium and then explains the core methods and approaches for analyzing distillation columns. Next, the author covers such topics as: Setting up a steady-state simulation Distillation economic optimization Steady-state calculations for control structure selection Control of petroleum fractionators Design and control of divided-wall columns Pressure-compensated temperature control in distillation columns Synthesizing four decades of research breakthroughs and practical applications in this dynamic field, Distillation Design and Control Using Aspen Simulation is a trusted reference that enables both students and experienced engineers to solve a broad range of challenging distillation problems.

It goes beyond traditional steady-state design that primarily considers the capital investment and energy costs when analyzing the control structure and the dynamic robustness of disturbances, and discusses how to maximize the economic and environmental benefits of reactive distillation technology.

Author : Alexandre C. Dimian,Costin S. Bildea,Anton A. Generic systematic methods are employed, supported by intensive use of computer simulation as a powerful tool for mastering the complexity of physical models. New to the second edition are chapters on product design and batch processes with applications in specialty chemicals, process intensification methods for designing compact equipment with high energetic efficiency, plantwide control for managing the key factors affecting the plant dynamics and operation, health, safety and environment issues, as well as sustainability analysis for achieving high environmental performance.

All chapters are completely rewritten or have been revised. This new edition is suitable as teaching material for Chemical Process and Product Design courses for graduate MSc students, being compatible with academic requirements world-wide. The inclusion of the newest design methods will be of great value to professional chemical engineers. Systematic approach to developing innovative and sustainable chemical processes Presents generic principles of process simulation for analysis, creation and assessment Emphasis on sustainable development for the future of process industries.

There are numerous books that focus on steady-state reactor design.